Convection

Convection is heat transfer associated with the movement of a gas or liquid. In a fire, hot smoke and heated air carry energy as they rise, move through openings or flow along surfaces. Convection also describes the resulting heat exchange between that moving fluid and walls, ceilings, structural members and other objects.

At a surface, convection combines fluid motion with molecular heat transfer through the adjacent fluid. Radiation can occur alongside it and is calculated separately. Natural movement caused by the fire and forced movement caused by wind or equipment can occur together.

Convection matters because it can transport heat far beyond the original fuel. It influences fire plumes, ceiling jets, smoke spread, structural heating and the activation of heat-responsive devices. The amount of energy transported and the heat delivered to a particular surface are related but distinct quantities.

Transport by moving gas and liquid

A flowing fluid carries its thermal energy with it. Hot combustion products leaving a flame therefore form a moving heat source, not only a cloud of combustion products. As this fluid travels, it can heat objects along its path and transfer energy into other rooms.

Two descriptions are needed for different engineering questions. The first follows energy carried by a volume or mass flow and gives a rate of energy transport. The second evaluates heat flux from fluid to a surface. Van Rossum notes that these approaches produce different result forms and should not be confused.

For energy carried by a flow, temperature, density, specific heat capacity and volumetric flow all matter. The temperature difference is taken against a defined reference, such as incoming ambient air. Changing any of these inputs alters the calculated energy transport.

For surface heating, the temperature difference between gas and surface is combined with a convective heat-transfer coefficient. That coefficient summarises the effect of fluid properties and flow conditions. It is not a universal constant for air, smoke and water, or for every orientation and velocity.

Free, forced and mixed convection

Free convection occurs when temperature differences create the flow. Hot combustion products have lower density than surrounding cooler air, so Buoyancy drives them upwards. Cooler air moves into the region and the process continues while heating and density differences persist.

Natural and forced convection use different driving influences. Based on Van Rossum (2027), §5.3.

Forced convection is imposed by an external cause such as a fan, pump, wind or mechanical smoke exhaust. These forces can strengthen, redirect or oppose the flow that the fire generates. The resulting pattern depends on their combined influence rather than on one label alone.

Mixed convection is common in buildings. A hot plume rises naturally while ordinary ventilation, wind at openings or a smoke-control system changes the surrounding pressure field. Even equipment not designed as fire equipment can affect heat and smoke transport.

Atreya further distinguishes flows inside bodies, such as pipes, from external flows around objects. Fire assessments commonly emphasise external natural convection around surfaces and within rooms. The chosen geometry determines which measured relationships and boundary conditions are appropriate.

Fire plumes and ceiling jets

A Fire plume forms as hot gases rise above a fire. Turbulent motion draws surrounding air into this rising stream. The plume's mass flow consequently increases while its temperature and combustion-product concentrations become more diluted.

Entrained air can greatly exceed the mass of burned fuel in the rising flow. The International Fire Engineering Guidelines therefore allow the fuel mass flow to be neglected in particular smoke-development methods. This does not mean the fuel is unimportant; its heat release establishes the buoyant flow that entrains the air.

A small fire can consequently produce a much larger flow of warm, diluted smoke. Convective heat transport is not proportional only to the visible flame volume. Air entrainment, plume height and surrounding geometry influence the quantity delivered upwards.

When the rising flow reaches a ceiling, it turns and spreads horizontally as a Ceiling jet. This hot, relatively fast-moving layer transfers heat to the ceiling and to objects near it. Flame-zone and ceiling-jet calculations inform evaluations of fire resistance, detectors and sprinklers.

A plume feeds a convective ceiling jet. Based on Van Rossum (2027), §5.3 and Australian Building Codes Board et al. (2005), p. 2.5-2.

As the ceiling jet moves outwards, entrained air dilutes it and heat transfer cools the gases near the ceiling. Accumulation of an upper hot layer changes the surrounding conditions. The initially unconfined ceiling-jet description therefore represents a particular stage of a compartment fire.

Coefficients, conservation and measurement

A convective Heat transfer coefficient has units of watts per square metre-kelvin when used for surface heat flux. Van Rossum also describes its reciprocal as a surface resistance. A larger coefficient means less resistance under the selected flow model, but it does not provide the total heat flow without area and temperature difference.

Surface heat flux is calculated by multiplying the gas-to-surface temperature difference by the convective coefficient. If the gas is hotter than the surface, the positive result describes heating of the surface. When the surface becomes hotter than the gas, the sign reverses and convection cools it.

$${\dot{q}}^{\prime }{\text{'}}_c=h_c(T_g-T_s),{\dot{Q}}_c={\dot{q}}^{\prime }{\text{'}}_{c}A$$

Symbol	Description	Unit
${\dot{q}}^{\prime }{\text{'}}_c$	Convective heat flux into the surface	W/m²
$h_c$	Convective heat-transfer coefficient	W/(m²·K)
$T_g,T_s$	Gas and surface temperatures	K or °C
$A$	Receiving surface area	m²
${\dot{Q}}_c$	Total convective heat-transfer rate for uniform flux	W

For a non-uniform exposure, flux can vary across the surface. The total transfer then requires adding the contributions over its area, or using an appropriate average coefficient. A local coefficient from one point does not automatically describe the entire object. Atreya distinguishes these local and area-averaged quantities explicitly.

Van Rossum gives indicative coefficients for several conditions. These are examples of the magnitude of surface transfer, not requirements or constants that apply to every fire. The applicable value depends on the flow at the surface. It also describes convection alone, so radiative exposure needs separate treatment.

Illustrative condition	$h_c$, W/(m²·K)
Ordinary indoor conditions	1–5
Smoke layer beneath a roof	3–10
Fully developed fire	20–25

The coefficient summarises processes within the boundary layer rather than a fixed property of the fluid alone. Surface geometry, flow velocity, turbulence and temperature-dependent properties can all influence it. Published coefficients and correlations therefore carry an application range.

Conservation laws provide the underlying framework. Atreya identifies conservation of mass, momentum and energy, supplemented by relationships for viscosity, conduction and diffusion. For turbulent flow, this framework often guides empirical correlations rather than yielding a simple universal solution.

Measurement must match the intended result. A gas-temperature reading alone does not determine the surface coefficient, and a surface heat-flux reading does not directly give the energy carried through an opening. Flow rate, density and reference temperature are needed for the transported-energy description.

The density calculation illustrates why units require attention. Van Rossum's exhaust example uses kelvin to determine smoke density, while the temperature difference between smoke and ambient air can be evaluated in degrees Celsius. Confusing these two uses of temperature would change the estimated energy transport.

Example of convective power carried by an exhaust flow. Based on Van Rossum (2027), §5.3.

Applying convective analysis

Convective transport can ignite light combustible materials at a considerable distance from the original fire. Van Rossum identifies this possibility when hot gases travel through buildings. The receiving material still needs a suitable time-dependent exposure and ignition condition; transport alone does not prove ignition.

Radiation can act alongside convection along the same path. A surface may receive radiation from a flame or hot gas layer while hot gas flows over it. Calculating only one contribution can underestimate exposure, while adding values based on different reference areas or times can create another error.

System operation changes the pressure field and therefore the convective path. Smoke exhaust can remove energy, ordinary ventilation can redistribute it, and wind can alter flow through openings. The analysis needs a defined system state rather than an assumption that all equipment either stops or operates as intended.

Plume models likewise apply to specified geometries. The 2005 guidelines describe axisymmetric models for fires away from walls, where air can be entrained from all sides. A wall, corner or obstruction changes the available entrainment paths and can place the fire outside the free-plume model assumptions.

Sensitivity to heat release rate, gas temperature and flow rate should be made visible. These quantities vary during fire growth and decay, so a single steady calculation cannot show when a detector, structural element or remote fuel package reaches a critical condition. Time selection is part of the engineering assumption.

Flow determination comes before surface heat transfer. Van Rossum emphasises that the gas movement must first be described, after which its thermal effect can be calculated. Assuming a heat-transfer coefficient before establishing whether the surface encounters plume, ceiling jet or slower room gas reverses that logic.

Natural and forced components can also change differently with time. A growing fire can strengthen buoyant flow while a fan operates at a fixed setting or a smoke-control system changes state. The combined flow therefore cannot always be scaled from fire size alone.

Results should be checked against their physical meaning. A surface heat flux needs a receiving area to become total heat flow, while transported power needs a defined flow crossing. Comparing these outcomes without conversion would mix local exposure with energy passing through the building.

Boundary layers and turbulence

Fluid immediately at a solid surface is approximately stationary, while velocity increases away from the surface. The resulting boundary layer contains strong velocity and temperature gradients. Heat crossing from the main gas flow to the surface must pass through this region.

In laminar flow, fluid moves in comparatively orderly layers with limited crosswise mixing. In turbulent flow, eddies transport fluid and energy across the main flow direction. Van Rossum explains that this stronger mixing generally increases convective heat transfer to a surface.

Turbulence is common near fires, openings and obstacles. Velocity, geometry and fluid properties all affect whether flow remains laminar or becomes turbulent. A coefficient obtained for one configuration cannot be transferred to another merely because both involve hot air.

The convective Heat flux to a surface depends on the gas-to-surface temperature difference and the heat-transfer coefficient. Higher gas temperature can increase this driving difference, while greater turbulence can increase the coefficient. Both effects must be considered when comparing exposures.

At the surface itself, molecular conduction remains part of the process even when the wider flow is turbulent. Convection therefore does not eliminate conduction; it changes the temperature and flow conditions under which that local transfer occurs.

Convection in smoke spread

Convection carries heated combustion products into upper parts of enclosures and through available openings. The resulting Smoke movement can expose locations remote from the fire. This remote exposure includes heat as well as reduced visibility and combustion-product concentration.

For a smoke flow, the transported power above an ambient reference follows from mass flow, specific heat and temperature rise. Multiplying density by volume flow first gives mass flow. The resulting power is carried through the exhaust or opening; it is not the heat delivered to any particular wall.

$${\dot{Q}}_{\mathrm{f}\mathrm{l}\mathrm{o}\mathrm{w}}=\rho c_p\dot{V}(T_x-T_0),\rho \approx \frac{353}{T_x}$$

Symbol	Description	Unit
${\dot{Q}}_{\mathrm{f}\mathrm{l}\mathrm{o}\mathrm{w}}$	Thermal power transported above the ambient reference	W
$\rho$	Gas density at smoke temperature	kg/m³
$c_p$	Specific heat at constant pressure	J/(kg·K)
$\dot{V}$	Volume flow at smoke temperature	m³/s
$T_x,T_0$	Smoke and reference ambient temperatures	K for density; K or °C for their difference

Van Rossum's worked example uses an exhaust flow of 100 m³/s at 463 °C, with ambient air at 20 °C. The source assumes air properties and a constant specific heat of 1005 J/(kg·K). Converting smoke temperature to 736.15 K gives a density of approximately 0.480 kg/m³. The density and volume flow must describe the same hot-gas state.

Retaining the unrounded density gives approximately 21.3 MW to three significant figures. This reproduces the source's approximately 21.35 MW with fewer reported digits. Rounding the density before multiplication slightly changes the result. The example illustrates sensible heat transport under fixed conditions, rather than an exhaust capacity suitable for every building.

$${\dot{Q}}_{\mathrm{f}\mathrm{l}\mathrm{o}\mathrm{w}}=\frac{353}{736.15}\times 1005\times 100\times (463-20)\approx 2.13\times 10^7\mathrm{W}=21.3\mathrm{M}\mathrm{W}$$

Wind and mechanical systems can modify or reverse expected paths. Natural pressure differences from heating occur together with expansion, stack effect and equipment-induced pressures. The whole pressure field must therefore be represented when assessing flow through openings.

Engineering use and limitations

Convection calculations support estimates of upper-layer energy, smoke production, boundary heating and device exposure. The required input can include heat release rate, plume entrainment, gas temperature and surface conditions. These inputs are often time-dependent and uncertain.

A steady surface calculation can provide an order-of-magnitude estimate, but a real fire changes both source and receiver temperatures. Assuming a fixed heat-transfer coefficient can hide these changes. A more detailed transient model is needed when the timing of exposure is central to the result.

Convective heat transfer is also not simply a direct result of air temperature. Two gas streams at the same temperature can deliver different heat fluxes because their velocities, turbulence and surface geometries differ. Conversely, a high coefficient cannot create heat transfer without a temperature difference.

Convection and smoke toxicity are separate aspects of the same transport. Dilution by entrained air can reduce concentration while increasing total mass flow. A diluted stream can still carry substantial energy and combustion products to another location.

The direction of heat exchange can reverse as surface and gas temperatures change. A construction that initially absorbs heat from smoke can later lose heat to cooler gas. Retaining the sign and time history of the temperature difference prevents this reversal from being lost in a single positive design number.

See also

• Adiabatic surface temperature
• Biot number

References

• Alpert, R. L. (2002). Ceiling jet flows. In P. J. DiNenno (Ed.), SFPE handbook of fire protection engineering (3rd ed., pp. 2-18–2-31). National Fire Protection Association.
• Atreya, A. (2002). Convection heat transfer. In P. J. DiNenno (Ed.), SFPE handbook of fire protection engineering (3rd ed., pp. 1-44–1-72). National Fire Protection Association.
• 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.
• Kandola, B. S. (2002). Introduction to mechanics of fluids. In P. J. DiNenno (Ed.), SFPE handbook of fire protection engineering (3rd ed., pp. 1-1–1-26). National Fire Protection Association.
• Rockett, J. A., & Milke, J. A. (2002). Conduction of heat in solids. In P. J. DiNenno (Ed.), SFPE handbook of fire protection engineering (3rd ed., pp. 1-27–1-43). 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., pp. 1-73–1-89). National Fire Protection Association.
• Van Rossum, E. (2027). Introduction to Fire Safety Engineering. Fire Safety Academy.