Carbon dioxide

Carbon dioxide (CO₂) is a gas produced by complete oxidation of carbon-containing fuel and is a normal product of combustion. In fire atmospheres it contributes to gas expansion, heat capacity and radiation, while elevated concentrations can stimulate breathing and cause asphyxiation. Its significance therefore extends beyond being a passive end product.

Complete Combustion converts fuel carbon to carbon dioxide and hydrogen to water, subject to the actual reaction and combustion efficiency. Incomplete combustion instead produces carbon monoxide, soot and other partially oxidised species. The relative yields depend on fuel, oxygen supply, mixing and temperature.

Carbon dioxide is also used as an extinguishing agent in some Gaseous extinguishing system designs, where it restricts oxygen available to the fire. The same property that suppresses combustion can make the atmosphere hazardous to people at sufficient concentration. Application therefore requires distinct fire-suppression and occupancy assessments.

Formation and fire chemistry

A balanced complete-combustion reaction specifies carbon dioxide and water among the stable products. Drysdale uses propane to show both the product quantities and the oxygen required for complete oxidation. The nitrogen in air accompanies these reactions and affects the gas mixture that must be heated.

The amount of carbon dioxide produced per unit fuel depends on fuel composition and the reaction pathway. Complete conversion is an ideal reference, while natural fires always include some partially oxidised products. Actual carbon dioxide yield therefore changes with material and combustion conditions.

Well-ventilated conditions favour more complete oxidation and higher carbon dioxide yield relative to carbon monoxide. Fuel-rich conditions reduce complete oxidation and increase carbon monoxide and other products. External burning can later oxidise these products to carbon dioxide after they leave the compartment.

Complete and incomplete combustion products. Based on Drysdale (2002), pp. 1-92–1-95.

Smouldering also produces carbon dioxide alongside carbon monoxide through direct surface oxidation of char and other condensed fuel. The ratio between them varies with material and airflow. Carbon dioxide production is therefore not limited to flaming combustion.

Physical role in fire atmospheres

Carbon dioxide contributes to the amount and heat capacity of combustion products. Energy released by combustion heats carbon dioxide, water vapour, nitrogen and other gases, so their thermal capacities influence flame and smoke temperatures. Drysdale lists carbon dioxide heat capacity at constant pressure as 37.129 J/mol·K at 298 K and 54.308 J/mol·K at 1000 K.

At high temperatures, carbon dioxide can dissociate as part of chemical-equilibrium changes. These heat-absorbing processes help make actual flame temperatures lower than simple adiabatic estimates. The gas is therefore chemically as well as thermally relevant to fire calculations.

Carbon dioxide also absorbs and emits thermal radiation in parts of the infrared spectrum. Tien and colleagues identify water vapour and carbon dioxide as important participating gases in the 1 to 100 micrometre range. This can affect radiative exchange between flames, smoke layers, boundaries and receiving objects.

Gas expansion during heating adds to pressure-driven movement through openings. Carbon dioxide forms part of this expanding mixture, although the resulting flow depends on the total gas amount, temperature and enclosure. It should not be treated as an independent driver isolated from other combustion products.

Physiological effects

Elevated carbon dioxide causes hypercapnia, an increased concentration in the body that alters breathing and can lead to incapacitation. Purser describes gradually increasing effects from about 3% to 6%, with more severe consequences above 7% and a growing risk of unconsciousness from 7% to 10%. These are approximate source-specific exposure relationships.

Purser also reports likely loss of consciousness within two minutes at 10% carbon dioxide in humans. This value must not be converted into a universal threshold for every exposure, but it demonstrates the severity of short high-concentration exposure. Concentration and duration still need to be assessed together.

Carbon dioxide stimulates ventilation, increasing breathing rate and volume. This can increase the uptake of carbon monoxide, hydrogen cyanide and lung irritants. The gas therefore affects both the person directly and the dose received from other smoke components.

CO₂ affects breathing and the uptake of other smoke products. Based on Purser (2002), pp. 2-108–2-110.

Oxygen depletion adds a separate hypoxic pathway. Carbon dioxide can accumulate while oxygen availability falls, and each change affects the fire atmosphere. Assessing carbon dioxide alone would omit the effects of reduced oxygen and the other combustion products.

Concentration and exposure time

Carbon dioxide does not follow Haber's simple constant-concentration-times-time rule. Purser contrasts approximate Ct products of 20%·min at 10% carbon dioxide and 175%·min at 5%, showing that concentration strongly affects time to incapacitation. A long low exposure and a short high exposure cannot be interchanged using one constant.

fractional effective dose (FED) for carbon dioxide therefore uses a concentration-dependent incapacitation relationship. Fractional doses are accumulated through time until the selected model predicts incapacitation. The method requires a stable or increasing concentration history, or another appropriate treatment of changing conditions.

Purser's worked example uses 5% carbon dioxide for 20 minutes followed by 9% for 2 minutes. The calculation below gives an accumulated fraction of about 1.01 at 22 minutes, corresponding to the model's severe-incapacitation endpoint. This is a worked application of the stated relationship, not a general exposure limit.

Worked carbon-dioxide exposure example with FED 1.01. Based on Purser (2002), p. 2-109.

Purser treats direct carbon-dioxide incapacitation separately from its enhancement of other-gas uptake. In his combined asphyxiant model, increased breathing modifies the carbon-monoxide and hydrogen-cyanide contributions, while oxygen depletion is evaluated through its own relationship. Adding a direct CO₂ dose indiscriminately to every other dose would not reproduce that model.

The timing of carbon dioxide exposure matters because fire growth changes both concentration and breathing demand. A person can initially inhale a dilute mixture and later encounter a much richer layer, while activity increases ventilation. Evaluating those periods separately is more defensible than averaging them into one exposure.

Different exposure metrics should remain attached to their purpose. Yield describes production, concentration describes the atmosphere at one place, and FED describes a modelled physiological outcome. Converting between them requires transport, breathing and dose-effect assumptions rather than a simple unit change.

Carbon dioxide measurements should also be interpreted with the oxygen measurement. Elevated carbon dioxide and low oxygen can occur together in a developed fire, and their physiological effects differ. Reporting only one concentration can make the atmosphere appear better characterised than it is.

For fire-suppression analysis, the relevant questions include initial concentration, retention time and leakage. A nominal agent quantity can fail to maintain the required atmosphere when openings and bypass flows continue. These same processes also influence the exposure of anyone remaining in or entering the protected space.

Model review should test the effect of external oxidation. Fuel-rich gases can leave the source room and produce carbon dioxide outside, changing both the source-room measurement and the external hazard. Assigning all of that carbon dioxide to the original compartment would give an inconsistent mass balance.

A traceable report therefore identifies the combustion mode, source yield, transport path, receiver concentration and exposure duration. It also distinguishes suppression performance from human tenability. These details allow carbon dioxide to be assessed as a combustion product, radiant gas, asphyxiant and extinguishing agent without confusing their different criteria.

Fire toxicity includes both asphyxiant and irritant effects. Carbon dioxide contributes primarily as an asphyxiant and breathing stimulant, while other products can injure airways, impair oxygen transport or prevent oxygen use in tissues. The combined mixture determines the exposed person's response.

Carbon monoxide reduces oxygen carriage, while hydrogen cyanide reduces the ability of tissues to use delivered oxygen. Carbon dioxide can increase the uptake of both and adds its own acidotic and asphyxiant effects. A hazard model based on one gas at a time can therefore underestimate the combined condition.

The complete exposure also includes heat and irritant smoke. These can impair vision and breathing and increase the time needed to escape. Longer exposure in a carbon dioxide-rich atmosphere then increases the accumulated dose.

Post-exposure effects matter as well. Purser describes evidence that carbon dioxide can enhance acidosis and lung injury after severe exposures in rodents. Such findings reinforce that removal from the atmosphere does not necessarily end physiological effects.

Calculating the direct carbon-dioxide contribution

Purser gives a concentration-dependent estimate of time to incapacitation by carbon dioxide. Concentration is entered as the numerical percentage by volume, so 5% is entered as 5. Each exposure interval contributes its duration divided by the incapacitation time at that interval's concentration. $$t_I(C)=\exp (6.1623-0.5189C),F_{\mathrm{C}{\mathrm{O}}_{\mathrm{2}}}=\sum _j\frac{\mathrm{\Delta }{t}_j}{t_I(C_j)}$$

Symbol	Description	Unit
$C_j$	Carbon-dioxide concentration during interval j	% by volume
$t_I$	Modelled time to incapacitation at a fixed concentration	min
$\mathrm{\Delta }{t}_j$	Duration of interval j	min
$F_{\mathrm{C}{\mathrm{O}}_{\mathrm{2}}}$	Accumulated direct carbon-dioxide dose fraction	dimensionless

For 5% carbon dioxide, the relationship gives approximately 35.4 min; for 9%, it gives approximately 4.45 min. Retaining unrounded intermediate values, the two intervals contribute 20/35.4385 and 2/4.44687. Their sum is approximately 1.01, matching the example reported by Purser.

The model reaches its reference endpoint when the accumulated fraction reaches one. That value describes the selected physiological endpoint; it is not a probability of injury and does not mean that lower fractions have no effect. Breathing changes and interaction with other fire gases can occur before direct carbon-dioxide incapacitation.

Purser specifies stable or increasing concentrations for this treatment. A decreasing exposure can involve recovery and cannot automatically be represented by the same simple accumulation. The relationship also concerns direct carbon-dioxide effects, so the calculation does not assess the complete toxicity of a fire atmosphere.

Use in fire suppression

Carbon dioxide can suppress fire by reducing oxygen available to the reaction zone. This is one of the traditional smothering mechanisms, distinct from cooling, fuel removal and chemical inhibition. Effective application requires the gas to reach and remain around the relevant fire region.

Deep-seated and porous fires require particular attention. In small-scale grain-silo tests described by Ohlemiller, carbon dioxide fed from the bottom was the most effective of the agents examined. That result belongs to the tested configuration and does not prove effectiveness for every enclosure or fire.

Hot fuel can reignite when oxygen returns before cooling is complete. Smothering therefore has to be maintained long enough to prevent renewed reaction, and concealed regions must actually receive the agent. Bypass channels can leave smouldering zones untreated.

Occupancy hazards remain even when suppression is successful. Elevated carbon dioxide can stimulate breathing and cause incapacitation, while oxygen depletion adds further risk. System evaluation must therefore distinguish the concentration required for suppression from conditions tolerable to people.

Measurement and modelling

A carbon dioxide yield gives the mass produced per unit fuel consumed and depends on combustion mode. Well-ventilated and vitiated test conditions can produce different yields, so the intended fire stage must be represented. Generic complete-combustion values are not a substitute for those data.

Source yield is not the same as concentration at a receiver. Plume entrainment, vent flow, mixing and external oxidation change the atmosphere during transport. A model must connect the source with the exposed location and the exposure duration.

Gas measurements also need appropriate units. Volume fraction, molar fraction and mass concentration require consistent treatment with the gas mixture and model. Ct products in Purser's examples use percentages by volume and time, which should not be confused with mass-based yields.

Radiation and temperature models need corresponding gas properties. Heat capacity and emissivity vary with temperature and wavelength, and simplified constant values apply only to their stated range. Source data should not be extrapolated silently to unrelated conditions.

Prevention and engineering use

Limiting ignition and fire growth reduces the total carbon dioxide produced along with other effluent. Smoke-control measures can then remove, dilute or contain the contaminated atmosphere. The resulting concentration and exposure duration at occupied locations must be evaluated.

Ventilation can support more complete combustion and reduce carbon monoxide while increasing carbon dioxide production and total fire growth. This illustrates why a lower carbon monoxide hazard does not automatically make the overall atmosphere safe. Fire development and gas composition must be considered together.

For suppression design, agent distribution and retention are as important as the nominal quantity. Openings and bypass paths can change concentration over time and allow reignition. The model should include these dynamics rather than assuming a single uniform fill.

Limits and common misunderstandings

Carbon dioxide is not harmless simply because it is a normal combustion product. High concentrations can cause rapid incapacitation, and moderate concentrations can increase uptake of other toxic products. Exposure conditions determine the hazard.

It is also not a direct measure of complete fire efficiency. Carbon dioxide forms in complete and incomplete combustion, and external oxidation can create it away from the original flame. Its presence alone does not establish that all fuel burned completely.

The gas is not merely an inert diluent in every fire model. It absorbs and emits radiation, affects heat capacity and stimulates breathing. These physical and physiological effects can be important alongside its quantity.

suppression with carbon dioxide does not guarantee permanent extinction or occupant safety. Reignition can follow oxygen readmission, and the resulting atmosphere can be dangerous. Both outcomes require separate evaluation.

This separation also helps prevent false reassurance. More complete oxidation can reduce carbon monoxide while producing carbon dioxide and heat, and successful smothering can create an atmosphere unsuitable for people. Each outcome needs its own evidence and acceptance conditions.

See also

• Carbon monoxide
• Hydrogen cyanide
• Tenability criteria

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.
• Grant, C. C. (2002). Halon design calculations. 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.
• Tien, C. L., Lee, K. Y., & Stretton, A. J. (2002). Radiation heat transfer. 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.