Flash point

Flash point is the lowest liquid temperature producing a momentary flash above its surface when a pilot flame is applied in a specified test. It describes a response of the vapour mixture, not the temperature at which the entire liquid must burn continuously.

The property is used as an indicator of how readily a liquid can produce an ignitable vapour mixture. A lower flash point generally corresponds to greater volatility under comparable conditions, but the measured value depends on the apparatus and procedure. Open-cup and closed-cup results are not automatically interchangeable. The reported method forms part of the result and must remain attached to the numerical value.

Flash point must be distinguished from fire point and autoignition temperature. At fire point, burning continues rather than flashing and ceasing. Autoignition occurs without a local pilot when the mixture is sufficiently hot. These distinctions make flash point useful for fire prevention while limiting its ability to describe a developed fire.

Vapour production and flammability

A liquid does not usually undergo flaming combustion throughout its bulk. Heat causes evaporation at the surface, and the vapour mixes with air above the liquid. Flaming occurs in this gas mixture when composition, temperature and ignition conditions are suitable. Sustained burning then depends on continued vapour production and heat feedback to the liquid.

Flammability limits define the fuel concentrations between which a premixed flame can propagate under stated conditions. The lower limit is the lean side, with relatively little fuel vapour, and the upper limit is the rich side. Flash point is related to the temperature needed to establish an ignitable mixture near the liquid surface, but it is a test property rather than a direct concentration measurement.

Beyler illustrates flammability with methane in air, which propagates flame between 5% and 15% methane by volume at normal temperature and pressure. A local pilot can cause combustion outside such limits without producing a flame that travels through the wider mixture. This distinction between local ignition and propagation is central to the flash-point test.

Vapour pressure links liquid temperature with the amount of vapour available, but the mixture above a real liquid is also affected by air movement and container geometry. Kanury notes that the vapour concentration above a vaporising surface depends on physical transport as well as fuel properties. Flash point therefore cannot be inferred from boiling point alone.

The surrounding pressure and oxygen concentration also matter. Flammability limits are specified at particular temperature and pressure, and changes in inert gas concentration alter the mixture. A tabulated flash point consequently describes standardised test conditions rather than every storage or process environment.

How the test works

In the definition used by Kanury, the liquid is presumed to have a uniform temperature. A small gas-flame pilot makes a quick sweep near the vaporising surface. The observed result is a momentary flame propagation through the vapour-air layer. Sustained burning is not required for the flash-point observation.

Flash-point tests use a pilot and a defined cup configuration. Based on Kanury (2002), pp. 2-190–2-191.

Closed-cup apparatus retains vapour more effectively and limits disturbance by room air currents. Open-cup apparatus exposes the sample more directly to its surroundings. Kanury reports that closed-cup methods, such as Pensky-Martens, generally produce flashing at a lower temperature than open-cup methods, such as Cleveland, in the cited comparisons.

The difference between methods is not always small or systematic. Table 2-8.1 lists naphtha at 34.4 °C in the Abel test and 50.0 °C in the Luchaire test. Kerosene is 52.8 °C by Abel and 60.0 °C by Cleveland. These examples demonstrate why a result needs its method, rather than serving as a universally interchangeable material constant.

Cup walls, airflow, heat transfer and vapour leakage all influence the observed flash. Closed cups reduce alterations caused by drafts, while still allowing limited air leakage into the test region. Conductive cup walls and the location of the pilot also affect the mixture and its heating. The measurement is therefore an operational property tied to a defined procedure.

Flash point, fire point and autoignition

At flash point, the applied pilot can produce a flame that flashes through the available mixture and then goes out. The vapour production rate is not yet sufficient for self-sustained burning under the test conditions. This is why the term should not be used as a synonym for the start of every liquid fire.

Flash point, fire point and autoignition describe different observations. Based on Kanury (2002), p. 2-190.

Fire point is defined similarly except that the flame must continue and sustain burning of the liquid. Continued heat feedback and vapour production are now sufficient. Table 2-8.2 gives open-cup flash/fire-point pairs of 92/97 °C for Texas solar oil and 133/164 °C for fuel oil. Their respective differences are 5 °C and 31 °C, demonstrating that no fixed temperature increment converts flash point to fire point.

Autoignition temperature is the minimum temperature at which a vapour or gas mixture with air ignites without a local pilot. The test can involve a heated flask, furnace, bomb, compression device or hot surface. Autoignition also involves ignition delay, so the observed response depends on exposure and method.

A liquid can have a low flash point and a much higher autoignition temperature because the two tests ask different questions. Flash point concerns the availability of an ignitable mixture for a pilot. Autoignition concerns whether the mixture can begin reacting without that pilot. Neither value alone describes all possible Ignition routes.

The 2005 International Fire Engineering Guidelines list acetone with a flash point of −18 °C and an autoignition temperature of 465 °C, and ethanol with corresponding values of 13 °C and 365 °C. Methanol is listed at 11 °C and 470 °C. These older tabulated values illustrate the distinction, but their source and test basis must accompany any application.

Flash point in liquid-fire development

A Pool fire requires continuing evaporation from the liquid surface and combustion of the resulting vapour. Heat from the flame returns to the liquid, supporting further vapour production. Flash point can indicate whether a piloted flash is possible at a lower liquid temperature, but it does not determine the later heat release rate or size of this developed fire.

Before a sustained pool fire develops, vapour may accumulate or disperse according to ventilation and container geometry. A spill in still air and a spill exposed to strong airflow can produce different local concentrations at the same liquid temperature. The relevant question is therefore whether an ignitable mixture occurs where an ignition source is present.

The 2005 guidelines explain that liquid fuels present a safety issue similar to gases when enough vapour enters the flammable range. Very low-energy ignition sources can then be sufficient. Their preventive strategy for gases and vapours is to keep concentrations below the lower flammable limit, rather than relying on a rich mixture remaining above the upper limit after dilution.

Flash point can support this assessment by showing how liquid temperature affects vapour availability under the test method. It does not measure vapour concentration at a particular distance from a spill, nor the ignition energy of that mixture. Those quantities require separate information about the scenario.

Use in prevention and classification

For Hazardous area classification, flash point supplies information about the liquid rather than the extent of a flammable atmosphere. The underlying ignition issue is whether released vapour mixes with air within its flammable range. This requires information about the release and mixing conditions as well as the liquid property.

For storage and handling, a lower flash point signals that an ignitable vapour mixture can develop at a lower liquid temperature under comparable conditions. Control can therefore address liquid temperature, vapour accumulation and ignition sources together. The fire-triangle elements remain separate even though flash point combines their effects in one test observation.

The property is also useful when comparing liquids through consistent data. Comparisons should use the same method and edition of the source, with units and any pressure conditions retained. Mixing an open-cup value for one liquid with a closed-cup value for another can distort the comparison.

Flash point is not a measure of toxic hazard, soot production or total chemical energy. It does not state how much fuel is present or how fast it will burn after sustained ignition. Those aspects need combustion, toxicity and fire-scenario data rather than a single ignition-related temperature.

Comparing liquids and source data

A meaningful comparison starts with values obtained on the same basis. The 2005 International Fire Engineering Guidelines place flash point beside flammability limits, autoignition temperature and boiling point for common gases and vapours. This arrangement helps distinguish the properties, but does not imply that one can be calculated reliably from another without a suitable model and conditions.

The table gives benzene a flash point of −11 °C, toluene 4 °C, styrene 31 °C and decane 46 °C. These are the temperatures reported by that compilation; the table does not identify the cup method for each entry. They do not rank the later heat release rates or total hazards of fires involving these liquids.

Selected flash points tabulated in the 2005 IFEG. Based on Australian Building Codes Board et al. (2005), pp. 3.3-2–3.3-3.

The same source lists lower and upper flammability limits alongside these temperatures. Benzene has listed limits of 1.3% and 7.9% by volume, while ethanol has 3.3% and 19%. Such data describe gas-mixture composition, whereas flash point describes liquid temperature in a particular apparatus. Keeping the quantities separate prevents a temperature from being mistaken for a vapour concentration.

Kanury's multi-method table provides a second kind of comparison. Gas oil is listed at 90.6 °C by Pensky-Martens, 93.3 °C by Cleveland and 92.2 °C by Luchaire. The pattern across methods is not sufficiently consistent to support one universal correction between all tests.

Heavier fuels tending to have higher flash points in these data is an observed trend, not a substitute for measurement. Kanury also notes that several trends across fuel families are weak or conceptually uncertain. A liquid's name or broad family therefore does not justify inventing a value when a tested result is unavailable.

Source age and scope also matter. The SFPE Handbook chapter dates from 2002 and the guidelines from 2005. Their values record the results assembled in those editions. In particular, the IFEG table does not identify a cup method for each flash point. Its entries cannot establish method-specific classification boundaries without that missing information.

Vapour concentration and a worked example

A simple evaporation model explains why flash point depends on both temperature and mixing. It first estimates the equilibrium vapour concentration immediately at the liquid surface. It then distinguishes that value from the concentration where the pilot passes. Kanury presents this approach as a physical explanation with significant predictive limitations.

At normal atmospheric pressure, the surface vapour fraction can be estimated from the liquid's normal boiling point and enthalpy of vaporisation. The latter is the energy needed to evaporate unit mass of liquid. The equation assumes ideal-gas behaviour, equilibrium at the surface and an approximately constant enthalpy of vaporisation. Temperatures in this equation are absolute temperatures in kelvin.

$$x_s=\frac{p_{v,s}}{p_0}=\exp \left[\frac{h_{fg}}{R{T}_b}(1-\frac{T_b}{T_s})\right]$$

Symbol	Description	Unit
$x_s$	Mole fraction of fuel vapour at the surface	Dimensionless
$p_{v,s}$	Equilibrium vapour pressure at the surface	Pa
$p_0$	Atmospheric pressure at the normal boiling point	Pa
$h_{fg}$	Enthalpy of vaporisation, approximated by its normal-boiling-point value	kJ/kg
$R$	Specific gas constant of the fuel vapour	kJ/(kg·K)
$T_b$	Normal boiling temperature	K
$T_s$	Liquid surface temperature	K

For ethanol, Kanury's table gives a normal boiling temperature of 351 K and an enthalpy of vaporisation of 837 kJ/kg. Its molecular mass is 46 kg/kmol. Dividing the universal gas constant, 8.314 kJ/(kmol·K), by that molecular mass gives the specific gas constant used below.

Take a surface temperature of 286 K, the closed-cup flash point listed in the same table. Substitution gives a surface vapour fraction of approximately 0.050, or 5.0% by volume. The unrounded values are retained during the calculation. This result estimates the surface concentration under the model assumptions; it does not independently predict the measured flash point.

$$R=\frac{8.314}{46}=0.180739\dots \mathrm{k}\mathrm{J}/\mathrm{(}\mathrm{k}\mathrm{g}\mathrm{K}\mathrm{)}$$ $$x_s=\exp \left[\frac{837}{(8.314/46)\times 351}(1-\frac{351}{286})\right]\approx 0.050$$

The same source lists ethanol's lean flammability limit as a volume fraction of 0.033. Above the surface, dilution and transport reduce the fuel concentration relative to its equilibrium surface value. The mixture at the pilot may therefore differ substantially from the calculated 5.0%. Its location and application time remain part of the experiment.

Kanury represents this reduction by a dimensionless transport factor smaller than unity. A propagating flash requires enough fuel at the pilot, together with an effective pilot and suitable thermal conditions. A model using only the surface value omits that transport problem. It also does not resolve the full energy balance or the empirical dependence of flammability limits on the apparatus.

These limitations explain why a boiling point cannot be converted directly into a reliable flash point. They also explain why a temperature below one reported cup-test result does not describe every possible release configuration. Flash point characterises the specified test, while fire development additionally depends on vapour supply, fuel inventory and heat feedback.

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.
• Beyler, C. L. (2002). Flammability limits of premixed and diffusion flames. In P. J. DiNenno (Ed.), SFPE handbook of fire protection engineering (3rd ed., pp. 2-172–2-187). National Fire Protection Association.
• Drysdale, D. D. (2002). Thermochemistry. In P. J. DiNenno (Ed.), SFPE handbook of fire protection engineering (3rd ed., pp. 1-90–1-98). National Fire Protection Association.
• Kanury, A. M. (2002). Ignition of liquid fuels. In P. J. DiNenno (Ed.), SFPE handbook of fire protection engineering (3rd ed., pp. 2-188–2-199). National Fire Protection Association.
• Van Rossum, E. (2027). Introduction to Fire Safety Engineering. Fire Safety Academy.