Zone model

A zone model is a fire model that represents a compartment using a small number of regions with uniform representative properties. A common arrangement divides the room into a hot upper gas layer and a cooler lower layer. The model calculates how their conditions change through fire, plume flow, ventilation and heat exchange. This simplification provides an efficient form of fire modelling, but it does not resolve every spatial variation within the compartment.

The hot gas layer is an idealised control volume with representative temperature, density and composition. Conservation relationships connect its changing state to the processes that add and remove mass and energy. Some zone approaches use one mixed region for appropriate conditions, while others retain a two-layer representation. Capabilities and transition assumptions differ between implementations, so the label alone does not define a complete calculation method.

The compartment idealisation

A local fire commonly produces a rising plume that supplies hot gases to the upper part of a compartment. Cooler gas remains below the developing layer. Van Rossum describes this as a stratified situation represented by a two-zone model. The boundary between the layers is a mathematical interface within that representation.

Each layer's assigned properties stand for conditions across its volume. A real fire can have temperature and concentration gradients both vertically and horizontally. The hottest location near the plume need not have the same conditions as a remote corner. Uniformity is a simplifying assumption, rather than a claim that measurements would be identical throughout the layer.

Zone models replace spatial variations with representative properties in a small number of regions. Based on Van Rossum (2027), §§9.1–9.2.

The lower layer is described as relatively cool. It should not automatically be interpreted as completely free of combustion products. Connected compartments and gas exchange can introduce species into either layer. A model that tracks composition must therefore be interpreted through its calculated concentrations, rather than the informal names given to the layers.

Some methods use a single mixed zone when the assumed conditions no longer support a distinct upper and lower region. Van Rossum describes possible transition conditions involving temperature, layer depth or fire extent. These choices belong to the particular model. A universal automatic switch at one fixed flashover threshold cannot be inferred for every zone program.

The number of zones is separate from the number of connected rooms. A building model can represent several compartments, each with its own simplified gas regions. Openings then connect their mass and energy balances. CFAST is an example of a dynamic model that represents such connected compartments.

Mass, energy and layer volume

Mass conservation accounts for all gas entering and leaving a layer. The plume can transfer gas from the lower layer to the upper layer. Ventilation can add or remove gas, and openings can connect the layer with another space. The difference between additions and removals determines the change in stored mass.

Mass conservation, energy conservation and the gas-state relationship jointly determine layer conditions. Based on Peacock et al. (2026), pp. 7–8.

For a short interval in which the chosen flow rates are held constant, the balance can be written as below. The second expression relates stored mass to density and volume. These relationships explain the accounting, but they do not independently determine density. Temperature, pressure and the energy balance are also needed in a complete model.

Δm = (ṁ_in − ṁ_out)Δt

m = ρV

Symbol	Description	SI unit
Δm	Change in layer gas mass during the interval	kg
ṁin	Sum of mass flows entering the layer	kg/s
ṁout	Sum of mass flows leaving the layer	kg/s
Δt	Time interval	s
m	Stored gas mass	kg
ρ	Representative layer density	kg/m³
V	Layer volume	m³

Energy conservation determines how stored energy changes through fire input, gas transport and heat exchange. The CFAST formulation also accounts for work associated with the changing layer volume. Together with an equation of state, these relationships connect mass, temperature, volume and pressure. Solving only the mass equation would leave the thermal state undetermined.

The mass stored in a smoke layer is not simply the mass of soot. It includes the gas mixture supplied by the plume and other flows. Entrainment adds surrounding air to the combustion products before they enter the layer. A calculation based solely on burned-fuel mass would therefore omit an important part of the gas supply.

Heating can change volume without a corresponding increase in mass. Cooling can likewise reduce the volume associated with a given mass at unchanged pressure. Consequently, interface movement reflects both storage and thermodynamic state. Treating the layer as an incompressible container filled at one fixed volume rate would remove that coupling.

For a rectangular room of constant plan area, layer volume determines layer depth directly. In more general geometry, the relationship between height and enclosed volume must be defined by the implementation. The model cannot recover geometry that has not been represented. Its interpretation must follow the actual spatial simplification used in the calculation.

Plumes, openings and connected spaces

The plume model supplies mass and energy to the upper layer. Its entrainment relationship depends on the represented fire and the distance over which the plume rises. This makes plume selection important to the resulting layer development. A conservation equation can be solved correctly while its plume source term remains unsuitable for the actual arrangement.

Vent flow connects layer conditions with pressure differences across openings. The location and dimensions of a door, window or vent influence which gas region participates in the exchange. The flow carries energy and species as well as mass. Its effects therefore extend beyond a change in layer height.

A network of connected compartments can transfer smoke away from the room containing the fire. Each receiving compartment develops its own state through the transported mass and energy. The connection pattern is relevant to a network model, while the zone idealisation defines the conditions assigned within each compartment. Both aspects influence the resulting building representation.

Mechanical ventilation capabilities depend on the particular program. The introductory simplifications discussed by Van Rossum do not imply that every zone model excludes forced flow. CFAST, for example, accepts prescribed volumetric supply and exhaust. Its supplied formulation does not model the complete HVAC system from fan characteristics and duct resistance.

This distinction matters when a specified flow is used as a boundary condition. The calculation assumes that the assigned ventilation service is delivered under the modelled circumstances. It does not necessarily establish the fan and network performance required to produce that flow. The physical system and the compartment calculation must use consistent assumptions.

Heat transfer and smoke composition

The hot gas exchanges heat with walls, the ceiling, the floor and objects. Radiation and convection influence the surface heat flux, while conduction affects the temperature within construction materials. These processes change the energy retained by the gases. Material properties and boundary assumptions consequently influence the predicted layer temperature.

The amount of wall surface exposed to the upper layer can change as its interface descends. The heat-transfer problem therefore evolves with the layer geometry. The wall temperature also changes as heat enters the construction. A boundary cannot generally be treated as a fixed energy sink independently of time and material properties.

The fire's heat release is another implementation-dependent input. Many calculations prescribe a time-varying fire history rather than predict its growth from first principles. Some models modify burning when oxygen becomes insufficient, while others retain the imposed history. Van Rossum explicitly identifies ventilation-controlled burning as a model-selection issue.

Smoke concentration requires additional information about combustion products. Van Rossum describes relating soot production to the amount of fuel burned through a soot yield. CFAST similarly transports species according to production and interlayer flows. An apparently reasonable temperature result does not establish the correctness of an independently assumed smoke yield.

Concentration also depends on the volume occupied by the species. Adding clean entrained air can dilute a product even while its total mass increases. Heating or cooling can change the corresponding volume. Species mass, total gas mass and optical smoke properties are therefore distinct outputs or intermediate quantities within the modelling chain.

Worked mass-balance illustration

Consider an illustrative rectangular compartment with plan area 200 m² and height 4 m. Its upper layer initially has depth 1 m and a selected density of 0.80 kg/m³. The initial layer volume is 200 m³. Multiplying volume by density gives an initial gas mass of 160 kg.

For a selected 20-second interval, hold total inflow at 10 kg/s and total outflow at 5 kg/s. The net accumulation is 5 kg/s. Over the interval, the layer therefore gains 100 kg, giving a final mass of 260 kg. These chosen flows demonstrate the balance rather than represent a particular plume or exhaust design.

A positive net mass input deepens the illustrative layer when density is held fixed. Based on Peacock et al. (2026), pp. 7–8.

If density is also held at 0.80 kg/m³ for this isolated comparison, the final layer volume is 325 m³. Dividing by the 200 m² plan area gives a depth of 1.625 m. In the selected 4 m high compartment, the interface is then 2.375 m above the floor. All three geometrical results follow from the stated constant-density assumption.

This is a frozen-input accounting example, not a complete fire simulation. It does not solve the energy balance or predict whether density remains constant. In a dynamic model, plume supply, opening flow and temperature can all change during the same interval. The example isolates why a positive net mass input deepens a layer when the other quantities are fixed.

Smoke filling calculations must connect that accounting to the actual source terms. A filling time cannot be inferred from room volume alone. The fire history, plume, ventilation and thermal losses influence the evolving layer. The supplied zone framework combines these processes to calculate the state over time.

Dynamic and stationary calculations

A dynamic zone calculation starts from initial conditions and advances the state in time. The fire can grow, openings or ventilation can change, and boundaries can heat. The resulting history describes how the modelled system responds to those prescribed and calculated changes. Initial conditions therefore form part of the scenario, not a disposable setup detail.

A stationary calculation instead seeks a balanced condition for a selected fire and set of boundaries. For example, a smoke-exhaust calculation may examine whether removal can balance the gas supplied to a layer. Such a result describes the assumed balance. It does not independently establish when that condition develops after ignition.

The two approaches answer different questions. A stationary interface can coexist with continuing inflow and outflow, while a transient interface records changing stored conditions. Neither description means that gas motion has stopped. Distinguishing equilibrium from absence of flow prevents misinterpretation of a constant calculated layer height.

An equilibrium mass balance also does not by itself establish an equilibrium temperature. Energy may continue to accumulate or transfer to heating boundaries. Both balances are required to describe a stationary thermal state. The state variables are linked, but agreement in one balance does not guarantee agreement in the others.

Applicability and spatial detail

Zone modelling is most persuasive when the compartment develops the stratification assumed by the representation. Van Rossum's introductory limitations include simple geometry, an undisturbed plume and a stable interface. Those conditions identify where the simplification is easier to defend. Strong local disturbances or complex flow paths require closer examination of the chosen implementation.

The CFAST technical guide states that its correlations generally arose from relatively uncluttered, flat-ceilinged, room-like compartments. It directs users to validation evidence for other aspect ratios and configurations. A program accepting a particular geometry does not prove the accuracy of its predictions there. Input flexibility and demonstrated applicability are different matters.

Computational fluid dynamics provides a more spatially resolved modelling approach. A need to examine local flow or temperature differences can therefore motivate consideration of a different representation. That decision should follow the assessment question and available evidence. Zone calculations remain useful where their averages and submodels adequately represent the required quantities.

Capabilities also differ within the zone-model family. Multiple fires, connected rooms, mechanical flows and oxygen limitation may be available in one implementation and absent in another. Broad textbook simplifications should not be treated as universal software restrictions. The selected model's documentation determines which mechanisms are actually represented.

Interpretation and supporting evidence

Validation should match the quantity and conditions of interest. A model may reproduce layer temperature differently from target heat flux or interface height. The CFAST guide describes comparisons through quantity-specific bias and spread, including experimental uncertainty. A general claim of successful validation is therefore less useful than the relevant comparison evidence.

The calculated interface is an idealised boundary between representative regions. It is not a complete description of a fluctuating mixing zone or every local smoke incursion. Likewise, an upper-layer average cannot locate all hot spots. Decisions requiring those details need an appropriate model or additional evidence.

Interpretation should preserve the fire inputs, geometry, opening conditions, thermal properties and model limitations. The resulting layer histories can then be related to the assessment's criteria. Numerical precision does not remove uncertainty in those assumptions. A useful zone calculation makes both its physical accounting and its simplifications visible.

See also

• Hand calculations
• MQH correlation

References

• Peacock, R. D., McGrattan, K. B., Forney, G. P., & Reneke, P. A. (2026). CFAST: Consolidated Fire and Smoke Transport (Version 7). Volume 1: Technical reference guide (NIST Technical Note 1889v1, revision 7.7.7, 27 May 2026). National Institute of Standards and Technology.
• Van Rossum, E. (2027). Introduction to Fire Safety Engineering.