Flashover criteria

Flashover criteria are measurable or calculated conditions used to identify or estimate the transition from a localised compartment fire to widespread surface involvement. Common indicators include upper-layer temperature, radiant heat flux at floor level and the heat release needed to overcome compartment heat losses. They provide operational descriptions of flashover, whose onset does not have one universally precise temperature or time.

A criterion requires its measurement location, units and physical meaning to be stated. Absolute gas temperature differs from temperature rise, and a model switching condition differs from observed ignition throughout a room. Engineering correlations also depend on ventilation, compartment geometry and sometimes the thermal properties of the boundaries. Their results estimate the conditions associated with transition; they do not establish that a particular fuel arrangement must follow a prescribed fire-development sequence.

Physical meaning of a flashover criterion

A compartment initially contains a local fire and an accumulating upper layer of hot gases. As the fire develops, radiation from those gases and heated surfaces can increase the heating of other combustible items. Additional pyrolysis and ignition then increase the fire's heat release. Flashover describes the rapid transition associated with widespread involvement, rather than simply the existence of a hot ceiling layer.

Different observations can identify different aspects of that transition. Flames emerging through an opening, ignition of indicators at floor level and high measured gas temperatures need not occur simultaneously. Walton and Thomas therefore describe flashover as an imprecisely defined transition. Although gas temperatures from 300 to 650 °C have been associated with its onset, the range of 500 to 600 °C is more widely used.

A criterion makes an analysis reproducible only when its definition is explicit. For temperature, the definition includes whether the quantity is an average within the hot gas layer or a reading near the ceiling. For radiation, it includes the receiving location and whether the specified exposure applies across the floor. A numerical result without these details can conceal differences between experiments or calculations.

Temperature, floor exposure and heat-release criteria describe related but distinct aspects of transition. Based on Iqbal and Salley (2004), pp. 13-3–13-6.

Flashover also differs from the later ventilation-controlled burning regime. After widespread involvement, internal oxygen supply may limit heat release while pyrolysis continues. Conversely, an enclosure with restricted ventilation may never reach the conditions required for widespread ignition. The possibility of transition therefore depends on the available fuel and ventilation together with heat transfer to the boundaries. It cannot be decided from fuel quantity alone.

Temperature and radiant exposure indicators

Temperature-based criteria usually represent observations from compartment fire tests or a selected condition in a simplified calculation. The frequently quoted 500 to 600 °C range concerns gas temperatures associated with transition. Some analytical methods instead specify a temperature rise above ambient. For example, the SFPE presentation of the MQH threshold uses a rise of 500 K from an ambient temperature of 295 K. Its corresponding absolute gas temperature is approximately 522 °C.

These definitions must remain separate when comparing results. Substituting an absolute temperature of 500 °C for a rise of 500 K changes the thermal condition. The distinction is particularly relevant when a model reports kelvin temperatures while a project criterion uses degrees Celsius. Differences between the two definitions should be resolved before evaluating whether a calculated time history crosses the chosen threshold.

A second commonly used indicator is radiant exposure near 20 kW/m² at floor level. Iqbal and Salley describe this as an exposure capable of producing piloted ignition in many ordinary combustible materials. It represents the thermal feedback associated with widespread involvement, rather than a universal ignition threshold for every material. Actual ignition depends on the receiving material and its condition.

The reviewed experiments illustrate that variability. Newspaper indicators and plywood did not ignite over identical flux ranges, and temperatures near the ceiling differed from averages over the upper room. A reported observation therefore carries the details of the indicator and measurement arrangement. Agreement with a nominal radiation criterion should be interpreted within those conditions, particularly when a calculation assumes a uniform gas layer.

Upper-layer temperature and floor radiation are related through heat transfer, but they are not interchangeable measurements. The radiation received by the floor depends on the emitting gases, heated boundaries and geometric relationship to the target. A temperature criterion is therefore a practical indicator of the feedback process. It does not prove that every exposed item experiences the same flux or ignites at the same instant.

Heat release correlations

Simplified correlations estimate the minimum fire power associated with flashover in a compartment. They express different approximations to the balance between heat input, losses through openings and heat transfer to enclosing surfaces. The ventilation factor, opening area multiplied by the square root of opening height, appears in each of the following expressions. The coefficients require the stated units and should not be transferred unchanged to another unit system.

The Babrauskas correlation uses opening geometry alone. The SFPE chapter presents a fitted expression based on 33 compartment tests, primarily with wood and polyurethane fuels.

Q̇_FO,B = 750 A_o √H_o

The MQH correlation includes the thermal interaction with the enclosure through an effective heat-transfer coefficient and the enclosing surface area. Its flashover expression follows from the selected upper-layer temperature rise:

Q̇_FO,M = 610 √[h_k A_T A_o √H_o]

Thomas represents heat loss with separate terms related to enclosure area and opening geometry. The published expression is:

Q̇_FO,T = 7.8 A_T + 378 A_o √H_o

Symbol	Meaning	Unit
Q̇FO	Estimated heat release rate associated with flashover	kW
Ao	Opening area	m²
Ho	Opening height	m
AT	Total enclosing surface area, excluding openings	m²
hk	Effective heat-transfer coefficient	kW/(m²·K)

The heat release rate in these expressions must be compatible with the compartment energy balance. It is the fire's thermal input to that calculation, rather than simply the stored energy in the fuel load. The correlations estimate a power threshold. A separate fire-growth history is needed to estimate when a developing fire might reach it, and the assumed growth must remain physically achievable under the ventilation conditions.

The distinction between a balance estimate and an empirical fit is visible in the Babrauskas method. An initial heat-loss approximation gives a coefficient of 600 multiplying the ventilation factor. Comparison with tests leads to the fitted coefficient of 750 used here. These values are different stages of the method's development, rather than alternative units for one identical equation.

The derivation also compares the threshold with an estimated stoichiometric heat release rate. The fitted flashover expression corresponds to approximately half that reference rate. The observed tests occupied a broader range around this proportion, which explains why the fitted line cannot identify every individual transition precisely. The stoichiometric comparison is part of the empirical framework and should not replace the actual oxygen balance of a particular fire.

The SFPE chapter also gives a coefficient of 620 for an alternative rearrangement of the MQH relation. It states that either 610 or 620 is acceptable within the accuracy of the expression. These versions should be identified in a calculation record, but their small numerical difference should not be mistaken for precise knowledge of the transition. The experimental scatter and model assumptions remain more substantial limitations.

Boundary properties and range of application

The heat needed to establish a hot layer depends partly on energy absorbed by the walls and ceiling. In the MQH approach, the effective heat-transfer coefficient changes with thermal penetration into the lining. Its selection requires thermal conductivity, density, heat capacity and thickness, together with the relevant time regime. A coefficient describing one lining or exposure duration is not automatically appropriate to another compartment.

The Babrauskas expression omits those lining properties and the enclosure surface area explicitly. Its underlying tests covered heat release rates from 11 to 3,840 kW. Their ventilation factors ranged from 0.03 to 7.51 m^5/2, and surface-area-to-ventilation-factor ratios ranged from 9 to 65 m^−1/2. Those ranges describe the supporting dataset and provide context for judging an application outside similar compartment conditions.

The simpler expression should consequently be understood as an empirical estimate. Its coefficient incorporates the circumstances of the tests instead of independently representing every thermal process. The MQH and Thomas forms expose more of the role of surface area, but they still simplify the enclosure energy balance. Adding variables does not make a correlation a detailed prediction of the ignition sequence across individual furnishings.

Iqbal and Salley emphasise the influence of both insufficient and extensive ventilation. Very restricted ventilation may prevent the fire from developing the required heat release. Extensive ventilation can increase cooling and require a larger fire to reach a critical upper-layer condition. The resulting relationship is more complex than assuming that closing an opening always increases, or always decreases, the likelihood of flashover.

The temperature correlations also have limitations during initial smoke filling and rapid transient growth. Their simplified balances presume conditions compatible with the experimental and analytical assumptions. Walton and Thomas caution against applications before established outflow or where changing conditions invalidate the approximation. A result should therefore be checked against the stage of compartment development represented by the calculation.

Worked comparison for a small compartment

Consider the SFPE example compartment measuring 3.0 m by 3.0 m in plan and 2.4 m high. Its door is 0.60 m wide and 1.80 m high. The opening area is 1.08 m², and the total enclosing area excluding the opening is 45.72 m². For the MQH comparison, the source example uses an effective heat-transfer coefficient of 0.030 kW/(m²·K). These are inputs for a correlation comparison, rather than a measured flashover event.

The opening height in this calculation is the door height, not the full compartment height. The ventilation factor is 1.449 m^5/2. Substitution in the Babrauskas relation gives approximately 1,087 kW. The MQH expression with coefficient 610 gives 860 kW, while the Thomas expression gives 904 kW. Repeating the MQH calculation with coefficient 620 gives 874 kW. Calculations retain unrounded intermediate values before reporting the rounded results.

Method	Calculated threshold	Principal inputs beyond opening geometry
Babrauskas	1,087 kW	Empirical coefficient
MQH, coefficient 610	860 kW	Enclosure area and effective heat transfer
MQH, coefficient 620	874 kW	Same inputs, alternative coefficient
Thomas	904 kW	Enclosure area

For this example, the enclosure-area-to-ventilation-factor ratio also lies within the stated range of the Babrauskas dataset. This geometric agreement supports a comparison with the published correlation, but it does not establish identical fuel arrangement or lining behaviour. The source example's specified heat-transfer coefficient remains essential to the MQH calculation even when the geometry is shared by all three methods.

The range demonstrates the influence of the selected approximation. It should not be presented as a probability interval or averaged to create an apparently more accurate threshold. Each result derives from a particular empirical or energy-balance representation. Selecting a method requires consideration of the enclosure, thermal boundaries and conditions underlying the analysis. The calculation record should retain that selection alongside its input data.

Different correlations give different thresholds for the same compartment. Based on Walton and Thomas (2002), pp. 3-185–3-186.

None of these results provides a time to flashover by itself. A time estimate requires the assumed fire growth to be compared with the selected threshold while checking ventilation and thermal conditions. The effective heat-transfer coefficient may itself depend on the time being evaluated. Treating the tabulated powers as universal room properties would discard that dependence and the physical assumptions behind the calculation.

Model transitions and reporting

Zone models may use explicit conditions to change from a stratified two-zone representation to a single mixed zone. Van Rossum describes possible conditions involving upper- or lower-zone temperature, smoke-layer depth and burning area. These conditions control the computational representation of the enclosure. They should be reported as model transition rules rather than silently substituted for an experimental definition of widespread surface ignition.

For example, a rule based on the hot layer occupying most of the enclosure concerns the validity of maintaining two separate zones. A rule based on high upper-zone temperature instead concerns intense thermal feedback. Both can justify changing the model state under their stated assumptions, but they measure different features of the fire. A predicted switch therefore needs interpretation before being labelled a physical flashover time.

A zone-model switch requires interpretation before it is reported as physical flashover. Based on Van Rossum (2027), §9.3.

A useful report identifies the criterion, source, units, measurement or calculation location, and selected correlation. It also records opening geometry, enclosure surface area and any lining parameters. If several indicators are available, their individual times or threshold values can be compared without requiring them to coincide. Differences can reveal the effect of measurement location, model structure or the chosen operational definition.

Flashover criteria do not establish safe conditions before transition. They describe a major stage of fire development, while hazardous smoke and heat can arise earlier. Nor do the historical correlations alone establish a current regulatory acceptance condition for a particular project. Their appropriate role is to support an explicitly defined engineering assessment of compartment fire development within the evidence and assumptions stated.

References

• Iqbal, N., & Salley, M. H. (2004). Fire dynamics tools (FDTs): Quantitative fire hazard analysis methods for the U.S. Nuclear Regulatory Commission fire protection inspection program (NUREG-1805). U.S. Nuclear Regulatory Commission.
• Van Rossum, E. (2027). Introduction to Fire Safety Engineering.
• Walton, W. D., & Thomas, P. H. (2002). Estimating temperatures in compartment fires. In P. J. DiNenno (Ed.), SFPE handbook of fire protection engineering (3rd ed., pp. 3-171–3-188). National Fire Protection Association.