Fire can absolutely burn underwater, and it does so more often than most people realize. The key is that water does not extinguish fire by being some kind of anti-fire substance; it works by cooling fuel below its ignition point and by blocking access to oxygen. Any combustion system that gets around those two obstacles can sustain a flame beneath the surface. Reactive metals, specially engineered thermite, leaking natural gas, and even industrial cutting torches all demonstrate that fire and water coexist under the right circumstances.
Why Water Usually Wins
Fire needs three ingredients working together: fuel, an oxidizer (usually oxygen from the air), and enough heat to keep the reaction going. Water attacks all three at once. It absorbs enormous amounts of heat as it evaporates, cooling fuel below the temperature needed for combustion. It physically separates fuel from atmospheric oxygen. And the steam it generates displaces the surrounding air, smothering the flame. For an ordinary campfire or house fire, that triple assault is devastatingly effective.
But notice the assumption baked into that explanation: it assumes the fire depends on atmospheric oxygen and cannot generate enough heat to overcome water’s cooling power. Strip away either assumption, and water stops being such a reliable extinguisher. That is exactly what happens with certain metals, chemical mixtures, and engineered materials that carry their own oxidizer or react with water itself.
Metals That Treat Water as an Oxidizer
Some metals are so chemically reactive that they do not need atmospheric oxygen at all. Instead, they rip oxygen atoms directly out of water molecules and burn using that oxygen as fuel for their own combustion. Aluminum and magnesium are the best-studied examples. In laboratory experiments, micron-sized powders of both metals were ignited and sustained combustion in an environment of pure water vapor at roughly 2,500 K, with researchers confirming that oxygen from the air played no measurable role in the reaction.1Combustion and Flame. Combustion of fine aluminum and magnesium powders in water At those temperatures, the metals essentially decompose water, liberating oxygen for their own use and releasing hydrogen gas as a byproduct.
This is not a quirk that only shows up in a lab. Anyone who has seen a magnesium flare dropped into a swimming pool has witnessed it firsthand. The metal burns with blinding intensity because the reaction is fiercely exothermic: once started, it produces more than enough heat to sustain itself despite the surrounding water. Trying to extinguish a magnesium fire with water actually makes things worse, because the water provides additional oxygen for the metal to consume. Fire departments train specifically on this: a metal fire calls for dry sand or a specialized Class D extinguisher, never water.
Alkali metals like sodium and potassium take this a step further. They react with water so violently that they do not just burn, they explode on contact. The mechanism is not simply hydrogen gas igniting, as chemistry textbooks long suggested. Research has shown that when an alkali metal contacts water, surface atoms shed their electrons within picoseconds. Those electrons shoot into the surrounding water, leaving behind a cluster of positively charged ions that repel each other so forcefully that the metal tears itself apart before the hydrogen even ignites. The hydrogen catches fire a moment later, adding a secondary fireball on top of the initial explosion. This is why dropping even a small chunk of sodium into a bucket of water produces a violent bang and a burst of flame.
Thermite Burns Through Almost Anything, Including Underwater
Thermite is a class of mixtures, typically powdered aluminum and iron oxide, that burn at extreme temperatures by carrying their own oxidizer within the mixture. Because the iron oxide supplies all the oxygen the aluminum needs, thermite does not require air and can burn underwater, in a vacuum, or in almost any other environment. The challenge for underwater use is not chemistry but engineering: keeping water from quenching the reaction before it reaches full temperature.
Researchers have developed nanothermite formulations specifically designed for underwater operation. One approach uses hydrophobic coatings on aluminum nanoparticles, wrapped in a fluoropolymer matrix, to repel water during the critical ignition phase. These tiny thermite sticks, less than a millimeter in diameter, demonstrated stable self-sustaining combustion underwater, burning at a rate of about 0.6 centimeters per second.2Chemical Engineering Journal. Underwater self-sustaining combustion and micro-propulsion properties of Al@FAS-17/PTFE-based direct-writing nanothermite That rate is slow compared to thermite burning in air, and the flame was weaker and cooler. Water’s cooling effect does not disappear; it just fails to stop the reaction entirely.
A separate line of research explored what happens when hydrophobic metal powders are enclosed in air-pocket structures called “bubble marbles,” essentially tiny gas envelopes that keep water at bay just long enough for ignition. Thermite enclosed in these structures was successfully ignited underwater using focused microwaves, demonstrating potential applications in heat generation and even underwater thrust.3Physical Review E. Insertion and confinement of hydrophobic metallic powder in water: the bubble-marble effect
Gas Leaks Beneath the Surface
Not all underwater fire involves exotic metals. When natural gas escapes from a subsea pipeline or a wellhead blowout, it rises as a plume of bubbles through the water column. If enough gas reaches the surface and encounters an ignition source, it catches fire at the water-air interface, producing the eerie spectacle of flames seemingly dancing on the ocean surface. The 2021 “eye of fire” in the Gulf of Mexico, caused by a subsea pipeline rupture, was a dramatic public example of this phenomenon.
The gas itself does not burn underwater in these cases. It burns at the surface because that is where it first contacts atmospheric oxygen. But the underwater portion of the system matters enormously. The pressure, depth, and flow rate of the leak determine how much gas reaches the surface, how concentrated it is when it arrives, and whether it can sustain a stable flame or just produce brief flickers. Research on natural gas released underwater found that the leakage flow rate is the key factor: higher release pressures and larger leak openings produce more gas at the surface, making stable burning much more likely.4Process Safety and Environmental Protection. Experimental research on natural gas leakage underwater and burning flame on the water surface A stable underwater plume structure also extends how long the surface flame persists, even after the leak rate fluctuates.5Proceedings of the Combustion Institute. Experimental study on offshore fires in cross air flow above water induced by the underwater release of natural gas
There is also a middle ground between “gas burns at the surface” and “fire burns fully submerged.” Experimental work on flammable gas leaking underwater has measured flame temperatures in the range of about 770 to 870 K at the leak point, which is considerably cooler than a typical open-air flame. The water vapor that diffuses from the surrounding liquid into the combustion zone suppresses radiation and lowers the overall flame temperature, making standard fire-plume models unreliable for predicting behavior in these scenarios.6Energy. Experimental study on fire thermal characteristics of flammable gases leakage underwater The fire is real, but it behaves differently than anything you would see on land.
Underwater Welding and Cutting
Perhaps the most routine example of fire burning underwater is also the least glamorous: industrial welding and cutting performed beneath the surface. Divers and remotely operated vehicles regularly use thermal cutting processes to dismantle offshore structures, repair ship hulls, and maintain subsea pipelines. These operations involve sustaining controlled combustion or electrical arcs in direct contact with seawater.
The techniques vary widely. A study evaluating cutting methods for decommissioning North Sea oil platforms tested five different underwater thermal systems, including oxy-hydrogen torches, oxy-petrol systems, and flux-cored arc cutting. Three of these manual methods were later adapted for robotic operation on remotely operated vehicles.7ASME Digital Collection. Remotely Operated Underwater Thermal Cutting Processes for the Decommissioning of Large North Sea Platforms Each approach solves the water problem differently. Oxy-fuel torches supply their own oxygen in a pressurized stream, creating a pocket of gas around the cutting point. Arc-based methods generate such extreme localized heat that they vaporize the surrounding water, forming a pulsating gas bubble that momentarily shields the arc from quenching.8PubMed Central. Metallurgy and Mechanism of Underwater Wet Cutting Using Oxidizing and Exothermic Flux-Cored Wires
What makes underwater welding fascinating from a combustion standpoint is that the fire does not burn freely. It exists in a constantly collapsing and reforming vapor envelope. The electrical arc superheats a tiny sphere of water into steam, the steam bubble expands, the arc burns inside it, the bubble collapses as the surrounding water rushes back in, and then the cycle repeats. This pulsating rhythm produces the characteristic crackling sound that divers hear during wet welding. The “fire” in this context is less a steady flame than a rapid series of micro-combustion events, each lasting a fraction of a second.
Supercritical Water and the Blurred Line Between Burning and Dissolving
At the far end of the temperature and pressure scale, water itself transforms into something that blurs the boundary between liquid and gas, and between dissolving and burning. When water is heated above 374°C and pressurized beyond about 22 megapascals, it enters a supercritical state where it behaves as neither a true liquid nor a true gas.9PubMed Central. Supercritical Water Oxidation as an Innovative Technology for PFAS Destruction In this state, organic compounds that would normally be insoluble in water suddenly dissolve readily, and oxidation reactions accelerate dramatically.
Supercritical water oxidation is used industrially to destroy hazardous waste, including notoriously persistent chemicals like PFAS (the “forever chemicals” found in firefighting foams and nonstick coatings). In one demonstration, dilute firefighting foam with a chemical oxygen demand of nearly 4,750 mg/L was treated in a supercritical water reactor and came out at just over 5 mg/L, representing almost complete destruction of the organic compounds.9PubMed Central. Supercritical Water Oxidation as an Innovative Technology for PFAS Destruction
Is this “fire”? Not in any visual sense. There is no visible flame. But the underlying chemistry is the same: organic molecules are being rapidly oxidized, releasing heat and producing carbon dioxide and water. It is combustion by another name, happening inside water that has been pushed into an exotic physical state. The point is worth noting because it illustrates that “burning” and “being underwater” are not as inherently opposed as everyday experience suggests. Under the right conditions, water can be the medium in which combustion occurs rather than the substance that prevents it.
Underwater Rocket Propulsion
Military engineers have taken the concept of underwater combustion and pushed it toward a genuinely counterintuitive application: using seawater itself as the oxidizer for rocket propulsion. Supercavitating torpedoes and underwater vehicles can travel at extraordinary speeds by riding inside a bubble of gas that forms around their nose cone. Maintaining that bubble requires thrust, and one approach generates that thrust by burning a metal-based solid propellant that reacts directly with the surrounding seawater.10Journal of the Korean Society of Combustion. Technical Trend of Supercavitating Rocket Propulsion System
The elegance of this design is that the vehicle does not need to carry an oxidizer at all. It scoops up seawater, channels it into a combustion chamber where it reacts with a metal fuel like aluminum or magnesium, and the resulting hot gas provides both thrust and the expanding gas envelope that reduces drag. The concept inverts the normal relationship: rather than water being the enemy of fire, it becomes part of the propulsion system’s fuel cycle. These systems are still largely in development and testing, but the underlying chemistry is the same metal-water reaction demonstrated in laboratory experiments with aluminum and magnesium powders.
Why You Cannot Put Out Every Fire With Water
The practical consequence of all this science shows up in fire safety, and it is a point that catches many people off guard. Water is the default tool for firefighting because it works brilliantly on the fires most people encounter: burning wood, paper, fabric, and similar organic fuels that depend on atmospheric oxygen. But for several categories of fire, water is not just ineffective but actively dangerous.
Grease fires are the most familiar example in domestic settings. Cooking oil burns at temperatures well above water’s boiling point, and throwing water on a grease fire causes the water to flash into steam, spraying burning oil in every direction. Metal fires are even more hazardous. As the laboratory research on aluminum and magnesium combustion confirms, these metals strip oxygen from water molecules and burn hotter in response to being doused.1Combustion and Flame. Combustion of fine aluminum and magnesium powders in water A machine shop fire involving magnesium shavings, or a lithium battery fire in an electric vehicle, can intensify sharply if someone reaches for a water extinguisher. Electrical fires add yet another category where water is dangerous for different reasons, mainly the conductivity risk.
Fire extinguisher classification systems exist precisely because of these distinctions. Class A extinguishers use water for ordinary combustibles. Class B is for flammable liquids. Class C covers electrical fires. And Class D extinguishers, which use dry powder agents like sodium chloride or copper-based compounds, are designed specifically for combustible metals. Knowing which class of fire you are facing is far more important than simply knowing where the nearest water source is.
How Deep Can Underwater Fire Go
Depth adds another variable that complicates underwater combustion. As you go deeper, water pressure increases by roughly one atmosphere for every ten meters. That rising pressure affects both the chemical reaction and the physical behavior of the flame. For gas leaks, higher pressure at the leak source means gas must fight harder to rise, and the bubbles that reach the surface are smaller and more dispersed. The surface flame, if it forms at all, may be weaker. But the relationship is not simple: higher pressure at the source also means the gas exits at a faster rate, which can compensate for depth effects and still produce a vigorous surface fire.4Process Safety and Environmental Protection. Experimental research on natural gas leakage underwater and burning flame on the water surface
For self-oxidizing reactions like thermite or metal-water combustion, depth is less of an obstacle because these reactions do not depend on atmospheric oxygen reaching the flame. In principle, thermite can burn at any depth as long as it can overcome the cooling effect of the surrounding water, which intensifies under pressure because water’s heat capacity and convection rates increase. The nanothermite formulations designed for underwater use showed notably lower burning rates and weaker flames even at atmospheric pressure, suggesting that extreme depth would further suppress performance.2Chemical Engineering Journal. Underwater self-sustaining combustion and micro-propulsion properties of Al@FAS-17/PTFE-based direct-writing nanothermite The same applies to underwater welding: as depth increases, the gas bubble surrounding the arc compresses, the arc voltage changes, and the metallurgy of the weld is affected. Most wet welding is performed at moderate depths for these practical reasons, though specialized techniques extend the working range.
Supercritical water oxidation, ironically, gets easier at depth. The high pressures found in deep ocean environments are closer to the 22 megapascal threshold needed for water to go supercritical. Volcanic hydrothermal vents on the ocean floor, where water temperature and pressure both soar, naturally produce conditions where rapid oxidation of minerals occurs, though calling that process “fire” stretches the word past its everyday meaning.
Historical Uses of Waterproof Fire
Humans figured out how to make fire survive contact with water long before anyone understood the chemistry involved. Greek fire, the Byzantine Empire’s secret weapon from the seventh century onward, was a liquid incendiary that reportedly burned on the surface of water and could not be extinguished by it. The exact recipe was a closely guarded state secret and has been lost, but historians and chemists have speculated for centuries about its composition. Leading candidates include mixtures of crude petroleum, pine resin, sulfur, quicklime, and possibly saltpeter. Quicklime reacts exothermically with water, which could explain why dousing the flames made them flare up rather than die.
Modern incendiary weapons extended this principle. White phosphorus ignites spontaneously when exposed to air and is difficult to extinguish with water because its ignition temperature is so low that it re-ignites as soon as the water evaporates. Military flares use magnesium-based compositions specifically because they produce light and heat that water cannot suppress. The thread connecting Greek fire to a nanothermite stick burning in a laboratory water tank is the same insight: carry your own oxidizer, burn hot enough to resist cooling, and water loses its power to stop you.