Smoke and heat exhaust ventilation - Revision history

Lisa: Add Intermediate SVG figures with Fire Safety Academy watermark at planned locations; preserve article text

2026-11-02T17:14:08Z

Add Intermediate SVG figures with Fire Safety Academy watermark at planned locations; preserve article text

← Older revision		Revision as of 18:14, 2 November 2026
Line 16:		Line 16:

	The layer grows unless smoke is removed. The book gives a simple example: if a plume supplies 10 kg/s to the layer and the exhaust removes 8 kg/s, the mass of the layer grows by 2 kg every second.<ref name="VR-3.7">Van Rossum (2027), §3.7.</ref> In Zukoski's experiments, a hood above a fire kept the smoke layer at a constant depth by extracting exactly the mass flow that the plume delivered at the layer interface. The book compares this with a SHEV system that holds a stationary smoke layer.<ref name="VR-6.7" />		The layer grows unless smoke is removed. The book gives a simple example: if a plume supplies 10 kg/s to the layer and the exhaust removes 8 kg/s, the mass of the layer grows by 2 kg every second.<ref name="VR-3.7">Van Rossum (2027), §3.7.</ref> In Zukoski's experiments, a hood above a fire kept the smoke layer at a constant depth by extracting exactly the mass flow that the plume delivered at the layer interface. The book compares this with a SHEV system that holds a stationary smoke layer.<ref name="VR-6.7" />

			[[File:fse-smoke-and-heat-exhaust-ventilation-v1.svg\|thumb\|upright=1.5\|alt=Hall cross-section: fire plume rising into a smoke layer under the roof, smoke leaving through roof vents, fresh air entering through low inlets.\|Principle of smoke and heat exhaust ventilation: the smoke removed through the roof vents balances the mass flow that the plume delivers to the layer, while replacement air enters low down. Based on Cooper (2002b), pp. 3-219 and 3-223, and Van Rossum (2027), §6.7.]]

	Ceilings are often divided into [[smoke reservoir\|smoke reservoirs]] by curtains hanging from the roof, so that the smoke stays in the curtained section above the fire. NFPA 204, as cited by Cooper, recommends that these curtains reach down at least one fifth of the height between the fire and the ceiling. The ceiling jet below the ceiling is about half that deep, and the layer must be thicker than the jet.<ref name="CO-234">Cooper (2002b), p. 3-234.</ref> If the smoke layer drops below the bottom of the curtains, smoke flows into the neighbouring sections.<ref name="CO-229">Cooper (2002b), p. 3-229.</ref>		Ceilings are often divided into [[smoke reservoir\|smoke reservoirs]] by curtains hanging from the roof, so that the smoke stays in the curtained section above the fire. NFPA 204, as cited by Cooper, recommends that these curtains reach down at least one fifth of the height between the fire and the ceiling. The ceiling jet below the ceiling is about half that deep, and the layer must be thicker than the jet.<ref name="CO-234">Cooper (2002b), p. 3-234.</ref> If the smoke layer drops below the bottom of the curtains, smoke flows into the neighbouring sections.<ref name="CO-229">Cooper (2002b), p. 3-229.</ref>
Line 25:		Line 27:

	Hotter smoke does not always vent better. The flow through a vent of given size reaches its maximum when the absolute temperature of the layer is twice that of the ambient air, about 586 K. Above 1.3 times the ambient temperature, a rise of about 90 K, the flow stays within 80 % of this maximum, but cooler layers give much lower flows.<ref name="CO-224" /> The effective size of an opening is smaller than its geometric size, because the flow contracts as it passes through. Cooper and the International Fire Engineering Guidelines use a coefficient of 0.61. Manufacturers give their vents a similar factor, for example 0.65, and the product is the [[aerodynamic free area]].<ref name="CO-222">Cooper (2002b), p. 3-222.</ref><ref name="IFEG-2.5-7" /><ref name="VR-6.8">Van Rossum (2027), §6.8.</ref>		Hotter smoke does not always vent better. The flow through a vent of given size reaches its maximum when the absolute temperature of the layer is twice that of the ambient air, about 586 K. Above 1.3 times the ambient temperature, a rise of about 90 K, the flow stays within 80 % of this maximum, but cooler layers give much lower flows.<ref name="CO-224" /> The effective size of an opening is smaller than its geometric size, because the flow contracts as it passes through. Cooper and the International Fire Engineering Guidelines use a coefficient of 0.61. Manufacturers give their vents a similar factor, for example 0.65, and the product is the [[aerodynamic free area]].<ref name="CO-222">Cooper (2002b), p. 3-222.</ref><ref name="IFEG-2.5-7" /><ref name="VR-6.8">Van Rossum (2027), §6.8.</ref>

			[[File:fse-smoke-and-heat-exhaust-ventilation-v3.svg\|thumb\|upright=1.5\|alt=Curve of relative vent mass flow against layer-to-ambient temperature ratio, rising steeply to a maximum of 1 at a ratio of 2 and falling slowly after.\|Relative mass flow through a natural roof vent as a function of the ratio of layer to ambient absolute temperature. Above a ratio of 1.3 the flow is within 80 % of its maximum. Based on Cooper (2002b), p. 3-224.]]

	Wind can help or hinder natural vents. Flat roofs tend to have negative wind pressures whatever the wind direction. If the open inlets are on the windward side, the raised pressure inside increases the flow through the roof vents. If they are on the leeward side, the lowered pressure inside can disrupt the venting completely and even reverse the flow.<ref name="CO-226" /> For a low, square building, identical inlets on all four sides would avoid the problem.<ref name="CO-226" /> In Dutch practice, a natural system is required to take its air supply from two façades that are at least 90° apart.<ref name="VR-6.6">Van Rossum (2027), §6.6.</ref>		Wind can help or hinder natural vents. Flat roofs tend to have negative wind pressures whatever the wind direction. If the open inlets are on the windward side, the raised pressure inside increases the flow through the roof vents. If they are on the leeward side, the lowered pressure inside can disrupt the venting completely and even reverse the flow.<ref name="CO-226" /> For a low, square building, identical inlets on all four sides would avoid the problem.<ref name="CO-226" /> In Dutch practice, a natural system is required to take its air supply from two façades that are at least 90° apart.<ref name="VR-6.6">Van Rossum (2027), §6.6.</ref>
Line 31:		Line 35:

	A single extraction point with too much capacity can draw clean air from below the layer as well as smoke. This effect, called [[plugholing]], wastes part of the capacity and is worse when the layer is thin.<ref name="VR-6.8" /><ref name="CO-228">Cooper (2002b), p. 3-228.</ref> It was first studied for natural vents, and later work showed that the same analysis applies to fans.<ref name="MI-305">Milke (2002b), p. 4-305.</ref> Remedies are more extraction points with less capacity each, a deeper smoke layer, or directing the suction horizontally rather than vertically.<ref name="VR-6.8" />		A single extraction point with too much capacity can draw clean air from below the layer as well as smoke. This effect, called [[plugholing]], wastes part of the capacity and is worse when the layer is thin.<ref name="VR-6.8" /><ref name="CO-228">Cooper (2002b), p. 3-228.</ref> It was first studied for natural vents, and later work showed that the same analysis applies to fans.<ref name="MI-305">Milke (2002b), p. 4-305.</ref> Remedies are more extraction points with less capacity each, a deeper smoke layer, or directing the suction horizontally rather than vertically.<ref name="VR-6.8" />

			[[File:fse-smoke-and-heat-exhaust-ventilation-v4.svg\|thumb\|upright=1.5\|alt=Left: one large extract point pulling a funnel of clean air through a thin smoke layer. Right: three smaller extract points removing only smoke.\|Plugholing: one extraction point with too much capacity draws clean air through the smoke layer (left); several smaller points remove only smoke (right). Based on Van Rossum (2027), §6.8 and Cooper (2002b), p. 3-228.]]

	== Design ==		== Design ==
Line 51:		Line 57:

	The book takes the rule for choosing a plume model from NEN 6093. The Thomas model is the more reliable choice when the mean flame length is shorter than the diameter of the fire; for longer flames, the Heskestad plume is preferred.<ref name="VR-6.7" /> Because the height appears to the power 1.5, a higher clear layer needs much more exhaust. In SHEV design, the plume mass flow is the smoke that must be removed to keep the layer at its design height.<ref name="VR-6.7" /><ref name="CO-236" />		The book takes the rule for choosing a plume model from NEN 6093. The Thomas model is the more reliable choice when the mean flame length is shorter than the diameter of the fire; for longer flames, the Heskestad plume is preferred.<ref name="VR-6.7" /> Because the height appears to the power 1.5, a higher clear layer needs much more exhaust. In SHEV design, the plume mass flow is the smoke that must be removed to keep the layer at its design height.<ref name="VR-6.7" /><ref name="CO-236" />

			[[File:fse-smoke-and-heat-exhaust-ventilation-v2.svg\|thumb\|upright=1.5\|alt=Curves of plume mass flow against clear height from 1 to 8 m for fire perimeters of 8, 12 and 16 m, rising steeply with height.\|Plume mass flow at the underside of the smoke layer according to the Thomas model, for three fire perimeters. Based on Van Rossum (2027), §6.7.]]

	=== Worked example ===		=== Worked example ===

Daan: New FSE wiki article: Smoke and heat exhaust ventilation (written by Claude Opus 5.5)

2026-11-01T22:40:38Z

New FSE wiki article: Smoke and heat exhaust ventilation (written by Claude Opus 5.5)

New page

'''Smoke and heat exhaust ventilation''' ('''SHEV''') is the removal of hot, smoky gases from the upper part of a space in a fire, combined with the supply of cool, clean air into its lower part. The smoke can be driven out by its own buoyancy through roof or high-level vents, by fans, or by a combination of both.<ref name="CO-219">Cooper (2002b), p. 3-219.</ref> Natural ventilation and extraction, the two forms of SHEV, are two of the five techniques of [[smoke control]] listed in the International Fire Engineering Guidelines, next to zone pressurisation, dilution and containment.<ref name="IFEG-2.5-7">National Research Council of Canada et al. (2005), p. 2.5-7.</ref>

The aim is to keep a layer of clear air under the smoke for long enough. Cooper names three objectives. SHEV helps people escape by limiting the spread of smoke into escape routes, helps firefighters to enter and see the seat of the fire, and reduces damage from smoke and hot gases.<ref name="CO-219" /> The principle is a mass balance. If the system removes as much smoke as the fire plume feeds into the layer, the layer stops descending.<ref name="VR-6.7">Van Rossum (2027), §6.7.</ref>

Natural systems are typical for single-storey halls such as warehouses and for atria, while fans are used where wind, cool smoke or the building layout make natural venting impractical.<ref name="CO-219" /><ref name="CO-236">Cooper (2002b), p. 3-236.</ref> How well SHEV works depends on enough replacement air, on the temperature of the smoke and on the wind.<ref name="CO-223">Cooper (2002b), p. 3-223.</ref><ref name="CO-224">Cooper (2002b), p. 3-224.</ref><ref name="CO-226">Cooper (2002b), p. 3-226.</ref>

== Purpose and applications ==
Venting may happen by chance, for example when smoke escapes through windows broken by the fire. A designed SHEV system instead uses dedicated openings or fans, such as ceiling vents that open automatically when a fusible link melts.<ref name="CO-219" /> Cooper describes two typical building types. The first is a large, open hall with ceiling vents, often divided by curtains hung from the ceiling into separate sections. The second is an atrium with shops at ground and mezzanine level that open onto it.<ref name="CO-219" /><ref name="CO-220">Cooper (2002b), p. 3-220.</ref>

Large-volume spaces need this approach because smoke spreads through them freely. Barriers and pressure differences, the methods used in compartmented buildings, are not practical there.<ref name="MI-292">Milke (2002b), p. 4-292.</ref> The high ceiling also has a benefit: smoke becomes diluted and cooled as it rises and spreads, which lowers the hazard it poses.<ref name="MI-292" />

The Australian verification method distinguishes two systems. Automatic smoke exhaust systems extract smoke from the hot layer near the roof with fans. Automatic smoke-and-heat vents work on the same principle but rely on the buoyancy of the hot gases.<ref name="AB-C4-128">Australian Building Codes Board (2020a), Data Sheet C4, p. 128.</ref> For both, an adequate supply of make-up air at low level is an integral part of the system. The objective is to keep tenable conditions in the space and escape routes long enough for evacuation and to support fire brigade intervention.<ref name="AB-C4-128" />

== How it works ==
A fire produces a rising plume of hot gases. As the plume rises, it draws in large amounts of surrounding air, a process called [[entrainment]]. Most of what is seen as smoke is therefore entrained air.<ref name="MI-292" /> Entrainment increases the mass flow in the plume and the amount of smoke, while it also cools and dilutes the combustion products.<ref name="MI-293">Milke (2002b), p. 4-293.</ref> The plume feeds a hot layer under the ceiling, and the higher the plume has to rise before reaching the layer, the more smoke it delivers.<ref name="VR-6.7" />

The layer grows unless smoke is removed. The book gives a simple example: if a plume supplies 10 kg/s to the layer and the exhaust removes 8 kg/s, the mass of the layer grows by 2 kg every second.<ref name="VR-3.7">Van Rossum (2027), §3.7.</ref> In Zukoski's experiments, a hood above a fire kept the smoke layer at a constant depth by extracting exactly the mass flow that the plume delivered at the layer interface. The book compares this with a SHEV system that holds a stationary smoke layer.<ref name="VR-6.7" />

Ceilings are often divided into [[smoke reservoir|smoke reservoirs]] by curtains hanging from the roof, so that the smoke stays in the curtained section above the fire. NFPA 204, as cited by Cooper, recommends that these curtains reach down at least one fifth of the height between the fire and the ceiling. The ceiling jet below the ceiling is about half that deep, and the layer must be thicker than the jet.<ref name="CO-234">Cooper (2002b), p. 3-234.</ref> If the smoke layer drops below the bottom of the curtains, smoke flows into the neighbouring sections.<ref name="CO-229">Cooper (2002b), p. 3-229.</ref>

For smoke to leave at the top, the same amount of air must enter at the bottom. Cooper's design method assumes large inlets near the floor, with a total area at least 1.5 to 2 times the area of the open ceiling vents.<ref name="CO-223" /> This [[replacement air]] should be clean, enter below the smoke layer and arrive slowly, at about 1 m/s near the plume. Fast inflow can deflect the plume, increase entrainment and even make the fire burn faster.<ref name="MI-306">Milke (2002b), p. 4-306.</ref> In mechanical systems the make-up air supply must also be smaller than the exhaust. Otherwise the space is pressurised and smoke is pushed through openings into adjacent spaces.<ref name="MI-306" />

== Natural and mechanical systems ==
Natural SHEV uses the pressure created by a hot layer. Inside the layer the pressure at the ceiling is higher than outside, so smoke flows out through open roof vents. The deeper and hotter the layer, the larger this pressure difference.<ref name="CO-223" /> Cooper gives an example of a 3 m deep layer that is 100 K warmer than ambient air at 293 K. The pressure difference is then 9.0 Pa, the outflow velocity 2.7 m/s and the mass flow 2.4 kg/s per square metre of vent area.<ref name="CO-224" />

Hotter smoke does not always vent better. The flow through a vent of given size reaches its maximum when the absolute temperature of the layer is twice that of the ambient air, about 586 K. Above 1.3 times the ambient temperature, a rise of about 90 K, the flow stays within 80 % of this maximum, but cooler layers give much lower flows.<ref name="CO-224" /> The effective size of an opening is smaller than its geometric size, because the flow contracts as it passes through. Cooper and the International Fire Engineering Guidelines use a coefficient of 0.61. Manufacturers give their vents a similar factor, for example 0.65, and the product is the [[aerodynamic free area]].<ref name="CO-222">Cooper (2002b), p. 3-222.</ref><ref name="IFEG-2.5-7" /><ref name="VR-6.8">Van Rossum (2027), §6.8.</ref>

Wind can help or hinder natural vents. Flat roofs tend to have negative wind pressures whatever the wind direction. If the open inlets are on the windward side, the raised pressure inside increases the flow through the roof vents. If they are on the leeward side, the lowered pressure inside can disrupt the venting completely and even reverse the flow.<ref name="CO-226" /> For a low, square building, identical inlets on all four sides would avoid the problem.<ref name="CO-226" /> In Dutch practice, a natural system is required to take its air supply from two façades that are at least 90° apart.<ref name="VR-6.6">Van Rossum (2027), §6.6.</ref>

Mechanical SHEV uses fans to extract the smoke. Cooper recommends it where wind could seriously reduce the efficiency of natural vents and where the smoke is too cool for natural venting to be economical. It also suits spaces where ceiling vents are impractical, such as intermediate floors of multi-storey buildings.<ref name="CO-235">Cooper (2002b), p. 3-235.</ref><ref name="CO-236" /> The design concept is the same as for natural vents, but the fans must be selected for the temperature of the smoke they extract. An alternative is to push air in with fans below the layer and let the smoke escape through natural vents, which keeps the fans out of the hot gases.<ref name="CO-236" />

A single extraction point with too much capacity can draw clean air from below the layer as well as smoke. This effect, called [[plugholing]], wastes part of the capacity and is worse when the layer is thin.<ref name="VR-6.8" /><ref name="CO-228">Cooper (2002b), p. 3-228.</ref> It was first studied for natural vents, and later work showed that the same analysis applies to fans.<ref name="MI-305">Milke (2002b), p. 4-305.</ref> Remedies are more extraction points with less capacity each, a deeper smoke layer, or directing the suction horizontally rather than vertically.<ref name="VR-6.8" />

== Design ==
Most designs are set up as a steady-state problem. The designer chooses a design fire, a minimum acceptable height of the smoke layer interface and the ceiling height, and then determines the vent area or fan capacity needed.<ref name="CO-234" /> The design fire must represent a realistic arrangement of combustibles in the building.<ref name="CO-234" /> Some Dutch methods describe fire growth by the time the fire takes to reach 2 m² rather than 1 MW; NEN 6093, the Dutch standard for SHEV systems, is an example.<ref name="VR-4.5">Van Rossum (2027), §4.5.</ref> Time-dependent problems with growing fires and vents opening one by one require a computer model such as LAVENT.<ref name="CO-234" />

The central quantity is the mass flow that the plume delivers at the underside of the smoke layer. For relatively large fires, the book uses the plume model of Thomas, in which the mass flow depends on the perimeter of the fire and the height the smoke rises:<ref name="VR-6.7" />

<math display="block">\dot{m}_\mathrm{p} = 0.188 \, W_\mathrm{f} \, z^{3/2}</math>

{| class="wikitable"
|+ Symbols
! Symbol !! Description !! Unit
|-
| <math>\dot{m}_\mathrm{p}</math> || mass flow in the plume at the underside of the smoke layer || kg/s
|-
| <math>W_\mathrm{f}</math> || perimeter of the fire || m
|-
| <math>z</math> || height from the floor to the underside of the smoke layer || m
|}

The book takes the rule for choosing a plume model from NEN 6093. The Thomas model is the more reliable choice when the mean flame length is shorter than the diameter of the fire; for longer flames, the Heskestad plume is preferred.<ref name="VR-6.7" /> Because the height appears to the power 1.5, a higher clear layer needs much more exhaust. In SHEV design, the plume mass flow is the smoke that must be removed to keep the layer at its design height.<ref name="VR-6.7" /><ref name="CO-236" />

=== Worked example ===
Cooper uses wood pallets in 3 m by 3 m arrays as an example of stored goods.<ref name="CO-220" /> Such a fire has a perimeter of 12 m. Using the book's flame length relation, the Thomas model applies as long as the fire stays below about 3.3 MW.<ref name="VR-6.7" /> Suppose the hall is 7 m high and the smoke layer must stay 3 m deep, so 4 m above the floor. The plume then delivers 0.188 × 12 × 4<sup>1.5</sup> ≈ 18.1 kg/s to the layer. With a clear height of 3 m it would be 11.7 kg/s, and with 5 m it would be 25.2 kg/s.<ref name="VR-6.7" />

If the 3 m deep layer were 100 K warmer than the outside air, Cooper's result of 2.4 kg/s per square metre gives the vent area. Removing 18.1 kg/s then needs about 7.6 m² of open roof vents.<ref name="CO-224" /> According to Cooper's assumption, the inlets near the floor must have at least 1.5 to 2 times this area, about 11 to 15 m².<ref name="CO-223" /> US vent units range from 1.5 m² to 4.7 m² each, so this means two to six vents.<ref name="CO-228" /> If heat losses kept the layer cooler than assumed, the flow per square metre would fall and the vents would have to be larger.<ref name="CO-224" />

The example is a simplification. Real layer temperatures depend on heat losses to the roof and walls, and the assumption of no heat loss can overestimate them in spaces lower than about 10 m.<ref name="MI-302">Milke (2002b), p. 4-302.</ref> In zone models, the extraction capacity should ideally be larger than the supply capacity in volume terms, and more so the more the fire heats the air.<ref name="VR-9.3a">Van Rossum (2027), §9.3 (How a zone model works).</ref>

== Interaction with sprinklers ==
Combining SHEV with sprinklers has long been controversial. Sprinkler spray cools the smoke, so less smoke leaves through the vents than designed. At the time of Cooper's review, no generally accepted guidelines existed for designing combined systems.<ref name="CO-236" />

This [[sprinkler-vent interaction]] can also work the other way. The Australian Building Codes Board notes that a smoke vent or exhaust system may remove hot gases before sprinklers activate, which may affect especially early suppression sprinklers.<ref name="AB-B1-21">Australian Building Codes Board (2020a), Data Sheet B1, p. 21.</ref> Sprinkler failure also matters for fan systems. If the sprinklers do not limit the fire to its design size, a smoke exhaust system that depends on them can be overwhelmed or its components overheated.<ref name="AB-C4-132">Australian Building Codes Board (2020a), Data Sheet C4, p. 132.</ref>

A review of 34 position papers, summarised by Cooper, reached the following conclusions. Venting has no negative effect on sprinkler performance and does not change the activation of early sprinklers or the total number that open. A fire directly below a vent may delay the first sprinklers slightly, without significant effect.<ref name="CO-237">Cooper (2002b), p. 3-237.</ref> Venting does limit the spread of smoke, improves visibility for occupants and firefighters and reduces damage. If the sprinklers fail, venting remains valuable for manual firefighting.<ref name="CO-237" /> The review recommends early, grouped opening of vents and placing curtains over aisles rather than over storage.<ref name="CO-237" />

== Reliability and limitations ==
For performance-based design in Australia, the Board uses a typical reliability of 80 % for both smoke exhaust systems and smoke-and-heat vents, with a range of 70 to 85 %.<ref name="AB-C4-134">Australian Building Codes Board (2020a), Data Sheet C4, p. 134.</ref> This is lower than the 90 to 95 % it uses for wet pipe sprinkler systems in general and residential buildings.<ref name="AB-C2-71">Australian Building Codes Board (2020a), Data Sheet C2, p. 71.</ref> For vents, this value was adopted from exhaust systems because no separate data were available.<ref name="AB-C4-140">Australian Building Codes Board (2020a), Data Sheet C4, p. 140.</ref> Smoke-and-heat vents are sensitive to external conditions, fire size and the buoyancy of the smoke, so their efficacy may be relatively low.<ref name="AB-C4-135">Australian Building Codes Board (2020a), Data Sheet C4, p. 135.</ref>

SHEV is not useful everywhere. Dutch guidance on car parks notes that smoke and heat extraction systems, mostly providing up to ten-fold ventilation, can also spread fire and smoke within the car park, depending on where they are placed. They are basically only effective for clearing smoke once the fire is out. Such systems have not been used since 2012 because their added value was found to be virtually non-existent.<ref name="VG-38">Van der Graaf et al. (2021/2023), p. 38.</ref> The same guidance still lists ventilation and smoke and heat removal among the facilities that can improve an interior attack by the fire service.<ref name="VG-50">Van der Graaf et al. (2021/2023), p. 50.</ref> Outlets must be placed away from windows, air intakes and escape routes, so that the smoke does not harm the surroundings.<ref name="VG-40">Van der Graaf et al. (2021/2023), p. 40.</ref>

A common misconception is that a SHEV system removes all smoke from a space. A successful design removes the hottest and most hazardous smoke and slows or stops the descent of the layer, for example above eye level; it does not clear the space.<ref name="CO-237" /> Another is that more exhaust always helps; plugholing shows that a single large outlet can waste its capacity.<ref name="VR-6.8" /> SHEV also interacts with ordinary building services. Ordinary ventilation systems can strongly influence smoke spread, so it can be difficult to decide what they should do when the fire alarm operates.<ref name="VR-6.6" />

== See also ==
* [[Warehouse fire]]
* [[Thomas plume]]
* [[Atrium smoke control]]

== Notes ==
<references />

== References ==
* Australian Building Codes Board. (2020a). ''Data sheets: Fire Safety Verification Method'' (Handbook annex, Print version 1.1).
* Cooper, L. Y. (2002b). Smoke and heat venting. In P. J. DiNenno, D. Drysdale, C. L. Beyler, W. D. Walton, R. L. P. Custer, J. R. Hall Jr., & J. M. Watts Jr. (Eds.), ''SFPE handbook of fire protection engineering'' (3rd ed., pp. 3-219–3-242). National Fire Protection Association; Society of Fire Protection Engineers.
* Milke, J. A. (2002b). Smoke management in covered malls and atria. In P. J. DiNenno, D. Drysdale, C. L. Beyler, W. D. Walton, R. L. P. Custer, J. R. Hall Jr., & J. M. Watts Jr. (Eds.), ''SFPE handbook of fire protection engineering'' (3rd ed., pp. 4-292–4-310). National Fire Protection Association; Society of Fire Protection Engineers.
* National Research Council of Canada, International Code Council, Department of Building and Housing, & Australian Building Codes Board. (2005). ''International fire engineering guidelines'' (2005 ed.). Australian Building Codes Board.
* Van der Graaf, J., Hessels, T., & Rosmuller, N. (2023). ''Fire safety of indoor car parks accommodating electrically powered vehicles''. Nederlands Instituut Publieke Veiligheid. (Original work published 2021)
* Van Rossum, E. (2027). ''Introduction to Fire Safety Engineering''. Fire Safety Academy.

[[Category:Smoke Control]]