Can Carbon Monoxide Explode?

Carbon monoxide can absolutely explode. It is a flammable gas with a wide explosive range in air, and under the right conditions it burns violently enough to produce devastating overpressure. Most people associate carbon monoxide strictly with poisoning, but the gas has been causing industrial explosions for well over a century. The chemistry behind those explosions turns out to be stranger than you might expect, with trace amounts of moisture playing a surprisingly important role.

A Flammable Gas With a Wide Danger Zone

Carbon monoxide burns in air according to a straightforward reaction: two molecules of CO combine with one molecule of oxygen to form two molecules of carbon dioxide, releasing heat. What makes CO particularly hazardous is its flammable range. In air at normal pressure and temperature, CO can ignite at concentrations as low as roughly 12.5% and as high as about 74% by volume. That is an enormous window compared to many other common fuels. Natural gas (mostly methane), for instance, has a flammable range of only about 5% to 15%. CO’s autoignition temperature sits around 609°C (about 1,128°F), meaning an external spark or flame is usually needed to set it off at room temperature, but hot surfaces in industrial equipment or fire-damaged structures can easily reach that threshold.

CO is classified explicitly as a flammable and explosive gas, and any accumulation near an open flame presents a real ignition risk.¹ Because the gas is colorless and odorless, a dangerous buildup can happen without anyone noticing until an ignition source triggers the explosion.

The Strange Role of Water Vapor

One of the more counterintuitive facts about CO explosions is that pure, bone-dry carbon monoxide is actually quite difficult to ignite. Researchers have long known that the explosive reaction between CO and oxygen depends heavily on the presence of even tiny amounts of hydrogen-containing molecules, particularly water vapor. Early studies noted that CO–oxygen mixtures exhibit distinct upper and lower pressure limits for explosion at a given temperature, and those limits shift depending on how much moisture is in the system.²

The reason comes down to chain-branching chemistry. The CO–oxygen reaction on its own proceeds sluggishly because it lacks the free radicals needed to sustain a chain reaction. Water vapor, even at concentrations as low as 20 parts per million, provides the hydrogen atoms that kick-start radical production. In simplified terms, an oxygen atom reacts with water to produce hydroxyl radicals, which then react with CO to form carbon dioxide and a hydrogen atom, and that hydrogen atom reacts with oxygen to make more hydroxyl radicals. The cycle feeds itself and the reaction accelerates explosively.³

Adding hydrogen, methane, or water to a CO–oxygen mixture favors the onset of explosion, while adding inert gases inhibits it.⁴ In those laboratory experiments, explosions were observed above about 900 K (roughly 627°C) over a range of pressures. This moisture dependence means that in the real world, where air is never truly dry, CO always has enough trace water to be dangerous. It also means that attempts to study CO combustion in ultra-pure laboratory conditions can underestimate the actual hazard.

Where CO Explosions Happen in Practice

Carbon monoxide is generated wherever combustion is incomplete: furnaces running rich, car engines in enclosed garages, structure fires starved of oxygen, coal mines after a methane ignition, and many industrial chemical processes. The explosion risk is greatest in confined or semi-confined spaces where the gas can accumulate to within its flammable range before encountering an ignition source.

Steel mills and blast furnaces are classic high-risk environments. Blast furnace gas is a low-energy fuel mixture containing CO, hydrogen, carbon dioxide, and nitrogen. Even though the CO concentration in blast furnace gas is relatively modest and the mixture is heavily diluted with inert gases, it can still explode under the right conditions. Research on blast furnace gas mixtures has shown that increasing the initial temperature broadens the flammable range, meaning hotter process conditions make an already dangerous gas mixture explosive over a wider set of concentrations.⁵ Workers in these facilities follow strict protocols for purging, venting, and monitoring gas lines precisely because a CO-rich leak in a hot environment can ignite with little warning.

Coal mines face a different version of the same problem. After an underground methane explosion, large quantities of CO spread through the mine’s ventilation network. The CO itself then becomes a secondary explosion hazard. Studies of gas diffusion patterns after mine explosions show that CO concentrations in intake airways can reach dangerously high levels within seconds, and that shutting down ventilation fans, which might seem prudent after an explosion, actually worsens the situation by reversing airflow and spreading CO into previously unaffected areas.⁶ Resuming ventilation quickly after the initial blast helps push CO out and reduce the risk of a secondary ignition.

CO Mixed With Other Fuels Makes Things Worse

Carbon monoxide rarely exists in isolation in real-world hazard scenarios. It typically shows up alongside methane, hydrogen, and other combustible gases, and those mixtures behave differently than any single component alone. Syngas, a widely used industrial fuel made primarily of hydrogen and CO, illustrates this well. The explosion characteristics of syngas change dramatically depending on the hydrogen-to-CO ratio. Higher hydrogen content means faster flame speeds and higher peak pressures because hydrogen is extremely reactive. Increasing the proportion of CO in syngas actually reduces the measured explosion pressure, partly because CO burns more slowly and partly because mixtures with more CO lose more heat to the walls of any enclosed space.⁷

That does not mean CO-heavy syngas is safe. Even at CO-to-hydrogen ratios as high as 95:5, the mixtures still explode. Adding carbon dioxide as a diluent does reduce peak explosion pressures, but it takes substantial dilution to make a meaningful difference. With 20% CO₂ added, peak pressures dropped by roughly 20–39% depending on the mixture’s fuel-to-air ratio and hydrogen content.⁸ That is a useful reduction for engineered combustion systems, but it falls far short of making the gas non-explosive.

When a CO-dominant gas mixture leaks into an environment already containing methane, the combined hazard can be greater than either gas alone. Research testing CO-rich gas mixtures added to methane-air atmospheres found that the added CO increased the maximum explosion pressure and flame speed of fuel-lean methane mixtures, effectively making a marginally dangerous atmosphere significantly more explosive.⁹ In other words, even relatively small CO leaks into a space that already has some methane floating around can push the situation past the tipping point.

Backdraft and Structure Fires

Firefighters deal with CO explosions in a form most people have seen dramatized in movies: backdraft. When a fire in a sealed room uses up most of the available oxygen, combustion slows or stops, but the hot fire continues to pyrolyze materials, producing large quantities of CO and other flammable gases. The room fills with superheated, fuel-rich gas that is too oxygen-starved to burn. The moment someone opens a door or window and lets fresh air rush in, the mixture hits its flammable range and ignites in a violent fireball that blows outward through the opening.

Reduced-scale experiments on backdraft have shown that the key factor determining whether it occurs is whether the concentration of unburned fuel in the compartment exceeds a critical threshold. In those experiments using methane as the representative unburned fuel, backdraft occurred when the mass fraction of unburned gas in the compartment exceeded about 9.8%.¹⁰ In real structure fires, the unburned gas is a complex mixture that includes substantial amounts of CO, and the phenomenon behaves similarly. The practical takeaway for firefighters is that any sealed, smoldering fire with blackened windows, pulsing smoke, or a rush of air being sucked inward at a crack is potentially loaded with CO-rich gas ready to explode the instant oxygen arrives.

Poison First, Bomb Second

The reason most people think of carbon monoxide only as a poison rather than an explosive is that lethal poisoning concentrations are far lower than explosive concentrations. A CO level of just 1% in air can be lethal to breathe.¹ That 1% is roughly 10,000 parts per million. The lower flammable limit sits at about 12.5%, or 125,000 ppm, which is more than twelve times the lethal breathing concentration. In most domestic and workplace scenarios, a CO leak will incapacitate or kill people long before the gas concentration reaches explosion territory.

This ordering of hazards has practical implications for gas detection. Household CO detectors are typically set to alarm at concentrations measured in tens to hundreds of ppm, because the health threat arrives at those levels. Industrial sensors in steelworks, refineries, and mines are calibrated differently, tracking not just the low-ppm health hazard but also watching for the much higher percentages that signal explosion risk. A survey of CO alarm activations in English homes found that about a third of incidents traced back to faulty gas appliances and roughly one in ten to misuse of cooking methods.¹¹ These domestic concentrations are dangerous for health but essentially never reach explosive levels. The explosion scenario is almost exclusively an industrial, mining, or structural-fire problem.

That said, there is one domestic scenario that occasionally blurs the line. An attached garage with a running car engine in a tightly sealed space can theoretically accumulate CO into the percent range, especially if the space is small and poorly ventilated. Suicide attempts using car exhaust in enclosed garages have, in rare cases, been followed by explosions when an ignition source was present. Modern catalytic converters have dramatically reduced CO output from vehicles, making this scenario much less likely than it was decades ago, but it is not impossible with older cars or malfunctioning exhaust systems.

Why Pure CO Burns Differently Than You Would Expect

If you have ever watched a gas stove ignite, you have seen a fuel catch fire almost instantly. CO does not always cooperate that easily. As discussed earlier, its ignition chemistry depends on chain-branching reactions that require trace hydrogen-containing species. This makes CO behave less like methane or propane and more like a fuel with a conditional personality. Under dry, low-pressure laboratory conditions, researchers had to push temperatures above 900 K and carefully control pressure ranges between roughly 15 and 100 torr to observe explosions in CO–oxygen mixtures with very low hydrogen content.⁴

In practical terms, this means the explosion hazard from CO is highly context-dependent. In a humid factory environment at atmospheric pressure, CO within its flammable range will ignite readily from a spark. In a hypothetically ultra-dry, low-pressure environment, the same concentration might not ignite at all. The real world is overwhelmingly on the humid side of this spectrum, so the practical hazard is very real, but the underlying chemistry is worth knowing for anyone working with CO in controlled environments like labs or specialty industrial processes. Engineers designing combustion chambers for syngas, for instance, have to account for the fact that the CO component’s burn rate shifts with humidity and hydrogen content, and that heat loss to chamber walls is proportionally larger for CO-rich mixtures than for hydrogen-rich ones.⁷

CO Explosions Beyond Earth

In a tangent that would have seemed absurd a few decades ago, astrophysicists now think about CO explosions on other planets. As the catalog of known exoplanets has grown, researchers have started modeling which atmospheric compositions are stable and which would destroy themselves through combustion or explosion. Carbon monoxide and oxygen can coexist in certain exoplanetary atmospheres, particularly at temperatures between about 600 and 800 K and pressures above 1 bar, where thermochemical processes favor CO production. If CO and oxygen both accumulate to sufficient levels, the atmosphere could reach a combustion-explosion threshold.¹²

There is a catch: oxygen only builds up to meaningful concentrations when the atmosphere is relatively hydrogen-poor. Hydrogen-rich atmospheres tend to scavenge free oxygen before it can accumulate alongside CO. So the scenario applies mainly to a subset of exoplanets with specific chemical compositions, not to gas giants drowning in hydrogen. The practical consequence for planetary science is that combustion chemistry places a ceiling on how much CO and oxygen can coexist in an atmosphere. If a planet’s atmospheric models show both gases at high concentrations, something is probably wrong with the model, because the mixture would have already blown itself apart. It is a strange application of the same chemistry that makes CO dangerous in a mine shaft, scaled up to an entire world.

Suppressing CO Explosions

Given the wide flammable range and the sensitivity to mixed fuels, preventing CO explosions in industrial settings relies on a few overlapping strategies. The most fundamental is ventilation: keeping CO concentrations well below the lower flammable limit by ensuring adequate airflow. In mining, this is the primary line of defense, and the evidence from post-explosion studies confirms that maintaining active ventilation, even after an initial blast, is critical to preventing secondary CO explosions.⁶

Where ventilation alone is not sufficient, inerting provides a second layer. Flooding a space with nitrogen or carbon dioxide displaces oxygen and pushes the atmosphere outside the flammable envelope. Research on syngas mixtures shows that CO₂ dilution meaningfully reduces explosion severity, though it has to be a substantial fraction of the total atmosphere to make a real difference.⁸ In blast furnace operations and syngas handling, piping systems are routinely purged with nitrogen before maintenance work, and continuous gas monitoring watches for leaks that could bring the local atmosphere into the danger zone.

Temperature management matters too. While higher temperatures widen the flammable range of CO-containing gas mixtures, they also increase the likelihood of autoignition without a discrete spark source.⁵ Keeping hot surfaces below the autoignition temperature of CO and eliminating potential spark sources through intrinsically safe electrical equipment are standard practices in facilities where CO is present. The combination of ventilation, inerting, temperature control, and ignition-source elimination is what stands between a functioning steel mill and a catastrophic explosion.

1
IOP Publishing. Carbon monoxide in the process of uncontrolled combustion – occurrence, hazards and first aid
2
Nature. Effect of Water Vapour on Explosions of Carbon Monoxide
3
Symposium (International) on Combustion. Ignition kinetics of the carbon monoxide-oxygen reaction
4
International Journal of Chemical Kinetics. Kinetics and mechanism of the explosive reaction of carbon monoxide and oxygen
5
Inżynieria Mineralna. Explosion Characteristics of Blast Furnace Gas
6
Fuel. Research on CO diffusion law and disaster control strategy after gas explosion in high-gas mines
7
Fuel. Pressure history in the explosion of moist syngas/air mixtures
8
International Journal of Hydrogen Energy. Effects of H2/CO ratio and CO2 dilution on the explosion behavior and flame evolution of syngas/air mixtures
9
Journal of Loss Prevention in the Process Industries. Effects of a carbon monoxide-dominant gas mixture on the explosion and flame propagation behaviors of methane in air
10
Elsevier. Critical condition of backdraft in compartment fires: a reduced-scale experimental study
11
PubMed Central. Indoor carbon monoxide: a case study in England for detection and interventions to reduce population exposure
12
The Astrophysical Journal. Limitation of Atmospheric Composition by Combustion–Explosion in Exoplanetary Atmospheres