Carbon monoxide

Carbon monoxide is a toxic, combustible gas produced when carbon-containing fuel is not fully oxidised to carbon dioxide. It is one of the main toxic products of fire and can be generated in especially high amounts during smouldering and ventilation-limited flaming. Its hazard depends on concentration, exposure duration, breathing response and the presence of other fire products.

Carbon monoxide forms part of Incomplete combustion, in which fuel carbon remains partially oxidised. Well-ventilated flaming fires generally produce less carbon monoxide than fuel-rich conditions, although yields still depend on fuel and combustion arrangement. Smouldering can also generate high levels despite its lower heat release rate.

The gas is dangerous both near the fire and after transport to other spaces. Gottuk and Lattimer report that nearly 75% of fatalities due to smoke inhalation in the cited work occurred in remote locations. Conditions in the fire compartment strongly influence the concentrations that reach those locations.

Formation during combustion

Complete oxidation of carbon-containing fuel produces carbon dioxide and water. Carbon monoxide appears when oxidation is incomplete, whether because oxygen is limited, mixing is poor, the reaction is quenched or the combustion mode favours surface oxidation. Actual yields vary with material and fire conditions.

Early overventilated compartment fires can approach open burning, with little carbon monoxide formation in the simplified description. As the plume enters the upper layer and openings restrict airflow, underventilated fuel-rich conditions develop. Products of incomplete combustion then become much more significant.

Ventilation condition changes carbon monoxide production. Based on Gottuk and Lattimer (2002), pp. 2-54–2-55.

The Equivalence ratio compares the available fuel-oxidiser relationship with the stoichiometric requirement for complete combustion. It helps distinguish overventilated from underventilated conditions, but local and compartment values can differ. A single room-average ratio does not describe every flame and layer region.

Smouldering provides a different pathway. Oxygen reacts directly with a condensed fuel surface, producing mainly carbon monoxide and carbon dioxide in proportions that depend on fuel and airflow. Ohlemiller reports examples of 6% to 7% carbon monoxide by mole in one foam experiment and 10% to 22% for cellulosic insulation under varying airflow. These are specific test results, not universal yields.

Carbon monoxide can subsequently oxidise to carbon dioxide. External burning of fuel-rich gases can therefore reduce the carbon monoxide and smoke leaving a compartment, while increasing heat release outside it. The location of combustion affects both toxicity and thermal exposure.

Toxic action and exposure

Carbon monoxide reduces the blood's ability to carry and deliver oxygen to tissues. This creates asphyxia, with confusion and possible loss of consciousness. Incapacitation can occur before death and can prevent a person from escaping, making the sublethal effects central to fire safety.

Fire toxicity concerns the adverse effects of combustion products on people. Carbon monoxide is a major contributor, but the same smoke can also contain hydrogen cyanide, low oxygen, carbon dioxide and irritants. These products can have additive or interacting effects, so assessing carbon monoxide alone can underestimate the total hazard.

Hydrogen cyanide is particularly relevant to nitrogen-containing fuels and to some smouldering polyurethane foams. Its concentration-response relationship differs from carbon monoxide, with rapid incapacitation possible at high concentrations. Purser's comparison shows why one concentration-time rule cannot be applied to every toxic gas.

Carbon dioxide can increase ventilation and thereby the uptake of carbon monoxide and other inhaled products. It also has its own asphyxiant effects at elevated concentrations. Low oxygen adds another pathway to hypoxia, so the complete atmosphere matters.

Concentration, time and dose

A concentration describes how much carbon monoxide is present in the gas at a particular time and place. A dose combines this exposure with time and breathing response. Toxic-effect prediction therefore requires a time history rather than a single peak concentration.

Gottuk and Lattimer state that concentrations as low as 4,000 ppm can be fatal in less than an hour, while full-scale compartment fires can reach several percent by volume. These observations demonstrate the potential severity but do not provide a universal lethal threshold for every person and exposure.

Purser's analysis of active monkeys gives approximately 26,600 ppm·min for 1,000 ppm, 28,097 ppm·min for 2,000 ppm, 26,868 ppm·min for 4,000 ppm and 26,086 ppm·min for 8,000 ppm as Ct products associated with incapacitation. The roughly constant Ct relationship applies to that model and endpoint, not to every human or fire atmosphere.

Longer exposures do not always follow a simple constant-concentration-times-time rule. Uptake depends on breathing rate and the difference between inhaled and body concentrations. Purser illustrates that the inhaled concentration needed for a lethal carboxyhemoglobin level at four hours differs from the proportion predicted by Haber's rule.

Activity changes breathing and therefore uptake. Carbon dioxide can also drive hyperventilation, increasing the dose of other inhaled products. Any dose assessment should consequently state the assumed person, activity and atmosphere rather than treating dose as only elapsed time.

Fractional effective dose

Fractional effective dose (FED) is a modelling method that divides the dose received during each time interval by the dose predicted to cause a selected effect. The fractions are summed until they reach one, when the model predicts incapacitation or death. Different toxic gases and endpoints require their own relationships.

For carbon monoxide, the Ct product can be nearly constant over particular ranges, making the method comparatively direct. For carbon dioxide, hydrogen cyanide and low oxygen, concentration has a stronger influence on the time-effect relationship. Their FED calculations therefore differ from a simple universal Ct constant.

An FED model can combine several asphyxiant gases and, where appropriate, sensory irritants and heat. Purser notes that some subpopulations may require a lower safety threshold, using 0.1 FED in the cited discussion. A value of one is therefore a model endpoint for the selected population, not a universal safe-unsafe boundary for everyone.

The method depends on accurate concentrations and yields over time. In one example, assuming early well-ventilated yields throughout produced an FED of only 0.1 after 200 seconds, while allowing yields to become vitiated predicted an FED of 1. This illustrates how incorrect combustion assumptions can severely underestimate toxic hazard.

Transport and smoke inhalation

Smoke inhalation exposes a person to carbon monoxide together with particles, other gases and heat. It can cause respiratory pain, confusion and loss of consciousness, with effects determined by the complete mixture. Carbon monoxide is often the principal asphyxiant, but not the only hazard.

Transport can separate carbon monoxide from some particles. Van Rossum notes that soot can settle while gases spread farther through a building, allowing high carbon monoxide concentrations even where visibility remains greater than 20 m. Good visibility therefore does not demonstrate a safe atmosphere.

External burning can oxidise carbon monoxide before it reaches adjacent spaces, but it does not guarantee complete removal. Underventilated compartments and external burning both change with time, so concentrations at a remote receiver need a transport and combustion model.

The source room remains important even when victims are elsewhere. Production in the compartment controls the maximum input to the building, while openings and mixing determine dilution. Assessment should follow the gas from production through transport to inhalation.

Carbon monoxide exposure connects source, transport and receiver. Based on Gottuk and Lattimer (2002), p. 2-54 and Purser (2002), p. 2-88.

Yields and measurement

A carbon monoxide yield expresses the mass of carbon monoxide produced per mass of fuel consumed. It must be measured under conditions that represent the intended combustion mode. Well-ventilated and vitiated yields can differ by more than an order of magnitude.

Small-scale tests can provide yields for selected equivalence ratios and materials, but full-scale fires add changing ventilation, geometry and flow. Purser cautions that full-scale equivalence ratios and mass-loss histories are needed to use those data appropriately. Test results should not be applied as universal material constants.

Yield history changes modelled incapacitation. Based on Purser (2002), pp. 2-96–2-97.

The 2005 guidelines identify carbon monoxide as the principal toxic species for many fires, while warning that atypical materials require other species such as hydrogen cyanide. They recommend using combustion-reaction analysis or suitable experimental data to determine source concentrations. This keeps the assessment tied to the actual fuel.

Measurements near the fire do not automatically represent remote exposure. Deposition, mixing and external oxidation can change the atmosphere along its path. Source yield, concentration and inhaled dose are different quantities and should be reported separately.

Prevention and engineering assessment

Prevention begins with limiting ignition and fire growth, because a smaller fire produces less effluent. Adequate ventilation can reduce fuel-rich combustion, but added airflow can also increase fire growth and transport. System effects must therefore be evaluated together rather than assuming ventilation is always beneficial.

Smoke-control measures can remove or contain contaminated gas, supporting extraction, dilution, pressurisation and containment. The 2005 guidelines identify these as broad smoke-management approaches. Their effectiveness depends on maintaining the intended pressure and flow paths during the fire.

Detection and alarm provide time for response before a toxic dose accumulates. Early action can change the exposure duration and the fire's later yields. The assessment should compare the time required for detection and escape with the time at which modelled incapacitation occurs.

Scenario selection should include developed and fuel-rich stages where relevant. A test based only on early well-ventilated combustion can miss the period producing most carbon monoxide. Sensitivity to yield, ventilation and occupant activity should be reported.

Limits and common misunderstandings

Carbon monoxide is not visible evidence of fire. A clear atmosphere can still contain a hazardous concentration, particularly after particles have deposited or become diluted. Visibility and toxicity must be assessed separately.

It is also not produced only by smouldering. Smouldering can generate high proportions, but ventilation-limited flaming is another major source. The combustion mode and oxygen supply determine the yield in both cases.

There is no single safe or lethal carbon monoxide concentration for every situation. Exposure duration, breathing rate, other gases and individual susceptibility affect the outcome. Numerical examples from animal models and fire tests must not be treated as universal human thresholds.

Finally, dilution does not necessarily remove risk. A larger smoke volume can contain lower concentrations while spreading toxic gas to more people. Assessment needs both the diluted concentration and the total transported mass over time.

Exposure assessment should also distinguish concentration measured on a dry or wet gas basis and the corresponding mixture composition. Species production models use these definitions when converting yields into atmospheric fractions. A consistent basis is needed before concentrations from tests and models are compared.

The timing of external oxidation can change the apparent trend. Carbon monoxide produced inside may be measured outside before it has fully burned, or after oxidation has reduced it. Without the measurement location and travel time, the same concentration can support different interpretations of the fire.

Carbon monoxide's combustibility adds another scenario dimension. Fuel-rich smoke can provide a flammable atmosphere when oxygen is restored, while carbon monoxide oxidation contributes to heat release. This links toxic exposure with smoke explosion and external burning rather than treating them as unrelated outcomes.

Response measures should be evaluated for their effect on total production as well as dilution. Suppression can reduce mass loss and later effluent, but incomplete combustion can increase yield per kilogram. A lower concentration at one moment does not prove a lower accumulated dose without the full history.

For comparison between tests, material form and ignition method should be recorded alongside yield. Small samples can respond differently from composite furniture, and pilot arrangement can change early combustion. These details help explain why one material name can have different reported carbon monoxide yields.

A final review should check whether the selected population and endpoint match the decision. Lethality, incapacitation and escape capability are different outcomes, and early effects can occur at much lower doses. The assessment should therefore include the period in which a person can still respond and leave.

A traceable carbon monoxide assessment therefore links the fuel and combustion mode to a time-dependent source term. It then follows transport and oxidation to the receiver, where breathing and exposure duration determine dose. This structure prevents a source yield from being mistaken for an inhaled exposure.

The final comparison should state which effect is being assessed and what evidence supports it. A lethal endpoint, an incapacitating endpoint and an escape criterion can require different dose-effect data and safety margins. Keeping them separate makes the result more useful and avoids overstating what a single concentration proves.

See also

• Haber's rule
• Multi-sensor detector
• Smouldering

References

• Australian Building Codes Board, National Research Council of Canada, International Code Council, & Department of Building and Housing New Zealand. (2005). International fire engineering guidelines (2005 ed.). Australian Building Codes Board.
• Drysdale, D. D. (2002). Thermochemistry. In P. J. DiNenno (Ed.), SFPE handbook of fire protection engineering (3rd ed.). National Fire Protection Association.
• Gottuk, D. T., & Lattimer, B. Y. (2002). Effect of combustion conditions on species production. In P. J. DiNenno (Ed.), SFPE handbook of fire protection engineering (3rd ed.). National Fire Protection Association.
• Kanury, A. M. (2002). Flaming ignition of solid fuels. In P. J. DiNenno (Ed.), SFPE handbook of fire protection engineering (3rd ed.). National Fire Protection Association.
• Ohlemiller, T. J. (2002). Smoldering combustion. In P. J. DiNenno (Ed.), SFPE handbook of fire protection engineering (3rd ed.). National Fire Protection Association.
• Purser, D. A. (2002). Toxicity assessment of combustion products. In P. J. DiNenno (Ed.), SFPE handbook of fire protection engineering (3rd ed.). National Fire Protection Association.
• Van Rossum, E. (2027). Introduction to Fire Safety Engineering. Fire Safety Academy.