Skip to content
Upcoming courses

Fire spread between buildings

Tools
From Fire Safety Academy Wiki
(Redirected from Horizontal fire spread)
Contents
  1. 1Mechanisms of spread
  2. 2Radiation from openings and the view factor
  3. 3Ignition of the exposed building
  4. 4Regulatory approaches
  5. 5Protection measures
  6. 6Limits and uncertainty
  7. 7See also
  8. 8Notes
  9. 9References

Fire spread between buildings is the spread of a fire from the building in which it starts to a neighbouring building or to another part of the same site.[1] It usually happens when a fully developed fire sends intense heat radiation through windows and other openings in an external wall. The radiation heats the facade of the building opposite until its surface ignites.[2] Flames projecting from windows, burning embers and falling debris can also carry the fire across.[3]

Building regulations deal with this hazard under the objective of protecting other property from damage caused by fire.[1][4] Engineers limit the radiation that reaches the boundary by controlling the distance to the boundary, the size of unprotected openings in the external wall and the fire resistance of the wall itself.[5] In New Zealand, for example, the received radiation may not exceed 30 kW/m² at the boundary and 16 kW/m² at 1 m beyond it.[1] New Zealand fire statistics for 2005 to 2010 suggest that about 3% of structure fires spread to adjacent property.[6]

Mechanisms of spread

The International Fire Engineering Guidelines (IFEG, 2005) name three ways in which fire can spread beyond the room of origin to an adjacent building or property boundary. These are burning embers and other debris, radiation, and direct flame contact.[3] The New Zealand commentary on the verification method C/VM2 states that the spread between neighbouring buildings is predominantly caused by radiation through openings in external walls.[2] For this reason, most design methods concentrate on the thermal radiation from windows, doors and other unprotected areas.[7]

Two buildings either side of a property boundary; radiation, window flames and embers travel from the burning building to the neighbouring facade.
Ways in which a fire spreads to a neighbouring building: radiation through openings, flames projecting from windows, and embers and debris. Based on National Research Council of Canada et al. (2005), p. 1.6-7, and Ministry of Business, Innovation and Employment (2025b), p. 43.

Radiation becomes important once the fire inside has reached flashover. For analysis, the IFEG suggests assuming that spread through an opening occurs at flashover, because a fire before flashover releases too little heat to spread through openings.[8] This assumption also applies to closed windows of ordinary glass, but not to toughened or wired glass.[8] Even without glass breakage, radiation can pass through the glass of fixed or closed windows and ignite material in an adjacent building.[3]

After flashover, flames often emerge from the windows of the burning room and rise along the outside of the building. This external flaming occurs because the burning extends beyond the room itself.[9] Experiments have measured heat fluxes as high as 200 kW/m² on the wall above such a window.[9] The limiting distance method used in New Zealand ignores such flame projection from openings, which is one reason why it may not be conservative in every case.[2] The spread of fire upwards over the outside of a single building is a different hazard, known as a facade fire. C/VM2 treats it in a separate design scenario for vertical fire spread.[10]

Firebrands are flaming or smouldering airborne objects that emerge from a fire and can sometimes ignite combustibles far away.[11] In wildland fires, wind can carry large embers great distances and start spot fires at other locations.[12] The IFEG notes that it is often difficult to establish how far a burning brand rising in the thermal plume may be carried by the wind.[13] Depending on the building and its contents, the distribution of flying brands and embers may need to be considered, particularly in areas prone to bushfires.[14]

Burning debris can also carry a fire over much larger distances. The fires in World Trade Center 7 on 11 September 2001 were ignited by debris from the collapse of the north tower, about 110 m to the south.[15] Fires started on at least 10 floors, and those on six floors grew and burned until the building collapsed.[15] Wind also plays a role in more ordinary cases. Its speed and direction affect how far flames project from windows and how much heat the room loses.[14]

Radiation from openings and the view factor

A radiation calculation for fire spread between buildings has two steps. The first step estimates the radiation emitted by each opening of the burning building. The second determines how much of this radiation reaches a point on the facade opposite.[16] The emitted radiation depends strongly on temperature, because the radiation from a hot surface or gas is proportional to the fourth power of its absolute temperature.[16]

Regulations often prescribe the emitted radiation directly, so that it does not have to be derived from a fire temperature. C/VM2 (second edition, 2025) uses 83 kW/m² for unsprinklered firecells with a fire load energy density of up to 400 MJ/m².[17] It uses 103 kW/m² for fire load densities of 400 to 800 MJ/m² and 144 kW/m² above 800 MJ/m².[17]

The fraction of the emitted radiation that arrives at the target is given by the view factor. The view factor is a measure of how much of the target's field of view is taken up by the radiating surface.[16] It depends on the area of the radiating surface, the distance to the target and the orientation of the two surfaces. A person standing close to a radiating surface of 5 m² may have a view factor of almost 100%, while the sun gives a very small view factor because of its great distance.[16] Introduction to Fire Safety Engineering names the calculation of fire spread to a neighbouring building as an example of its use.[16]

The received radiation then follows from one simple multiplication. The view factor contains all the geometry, so the same factor applies to any value of the emitted radiation. The radiation arriving at the target equals the view factor times the radiation emitted per unit area of the opening:[16] q˙′'r=F⋅E

Symbol Description Unit
q˙′'r radiant heat flux received at the target point kW/m²
F view factor from the radiating openings to the target point –
E radiation emitted per unit area by the openings kW/m²

The relation assumes that the opening radiates uniformly over its whole area and that the target sees it directly. In the simple table and formula methods of the book, the radiating rectangle and the target surface are parallel, and the target lies opposite a corner of the rectangle.[16] Any other position is handled by adding or subtracting rectangles. Gases between the surfaces can absorb part of the radiation, which the relation ignores.[16] C/VM2 sets the emissivity of the fire gases to 1.0, which treats the opening as a perfect black radiator.[17][16]

Introduction to Fire Safety Engineering gives a worked example with two windows in one fire compartment. Each window is 1.8 m wide and 1.2 m high, the windows are 2.4 m apart and the facade opposite is 6 m away.[16] The openings emit 120 kW/m², and the limit for the received radiation in the example is 15 kW/m². The highest radiation on the opposite facade usually occurs on the line of symmetry between the two windows, not directly in front of one of them.[16]

Because of this symmetry, one quarter of the situation is calculated and multiplied by four. The view factor of an imaginary rectangle from the outer edge of a window to the line of symmetry is 0.014. The non-radiating wall between the windows accounts for 0.006, so the radiating quarter gives 0.008.[16]

Elevation of two windows with the target point on the symmetry axis; one quarter is shaded as a large rectangle minus the wall strip.
View factor of two windows to a point on the symmetry axis 6 m away: one quarter equals rectangle AEFG minus the wall strip KEFH. Based on Van Rossum (2027), §5.4.

The total view factor is therefore 4 × 0.008 = 0.032, and the received radiation is 0.032 × 120 = 3.84 kW/m².[16] This is well below the limit of 15 kW/m², so no further measures at the openings are needed. If the result exceeded the limit, the openings would have to be made fire-resistant or protected in another way.[16] The exact view factor formula gives 0.030 and thus 3.6 kW/m². The book's table method, which uses the next higher table value, is slightly on the safe side.

Ignition of the exposed building

Whether the received radiation sets the neighbouring building alight depends on the material of its facade and on the duration of exposure. The minimum heat flux needed to ignite a material is its critical heat flux for ignition. For wood heated for less than 30 minutes, the IFEG gives 12 kW/m² with a pilot flame or spark and 20 kW/m² without one.[18] For wood heated for several hours, both values fall to 4.3 kW/m².[18] For separation distances, the IFEG refers to an SFPE engineering guide on the piloted ignition of solid materials under radiant exposure.[19] Roof coverings can also be tested for ignition by radiation from an adjacent building combined with piloted ignition from flying brands.[13]

The New Zealand limits are based on this kind of reasoning. The limiting distance method assumes cladding properties that represent timber with a moisture content of 15%.[2] Ordinary timber cladding without fire-retardant treatment is expected to ignite in less than 15 minutes when exposed to 30 kW/m² in the cone calorimeter test ISO 5660.[6] Brick, concrete and steel are non-combustible.[6] The limit of 30 kW/m² at the boundary therefore only works together with stricter requirements for walls close to the boundary.[2]

Under the New Zealand Building Code, external walls closer than 1 m to the boundary must be non-combustible or must resist ignition under a radiant flux of 30 kW/m².[1] Buildings in importance levels 3 and 4 must not ignite for 30 minutes, and buildings in importance levels 1 and 2 for 15 minutes.[1] Such walls are expected to withstand 30 kW/m² at least until the fire service arrives.[2] Further from the boundary, the materials of the external wall are not restricted. The limit there is set lower, at 16 kW/m², but this value may not prevent ignition or damage to all cladding materials.[6]

Regulatory approaches

In New Zealand, the verification method C/VM2 (second edition, 2025) contains a design scenario for horizontal fire spread. In this scenario, a fully developed fire in a building exposes the external walls of a neighbouring building or firecell.[1] The Building Code limits the received radiation to 30 kW/m² at the relevant boundary and to 16 kW/m² at 1 m beyond it.[1] The relevant boundary is the boundary with other property; where the other property is a road, it is the far side of that road.[20] Between two buildings on the same property, a notional boundary is assumed instead.[21]

The calculation covers the radiation from all unprotected areas of the external wall. Unprotected areas include walls without a fire resistance rating as well as unrated windows, doors and other openings.[17] Designers may use tables of permitted unprotected area or a direct calculation. The calculation must account for the distance to the boundary, the size and shape of the unprotected areas and the prescribed emitted radiation.[7] Only one fire-separated space needs to be considered at a time as a source of radiation.[7] Unprotected areas are not allowed within 1.0 m of the boundary, apart from small openings and fire-resisting glazing.[7]

Line chart: radiation received in front of a 1.8 m by 1.2 m opening falls from over 50 kW/m² at 0.5 m to below 3 kW/m² at 6 m.
Radiation received opposite the centre of one 1.8 m × 1.2 m opening for the three emitted fluxes of C/VM2, with the limits of 30 and 16 kW/m². Based on Van Rossum (2027), §5.4, and Ministry of Business, Innovation and Employment (2025b), pp. 43–44.

The emitted radiation in C/VM2 is meant to match fire gases at the temperatures reached in a standard fire resistance test after 30, 45 or 90 minutes, depending on the fire load.[6] With an emissivity of 1.0, the value of 83 kW/m² corresponds to a radiating temperature of about 827 °C.[16][17] The design fire also assumes that firefighters will intervene after some time to prevent spread to neighbouring property.[17] The commentary describes the approach as the limiting distance method, which also underlies the acceptable solutions C/AS1 to C/AS7.[2]

In Australia, the Fire Safety Verification Method (2022) contains a similar design scenario. It requires that the risk of fire spread between buildings is not greater than for a building that complies with the deemed-to-satisfy provisions.[22] The scenario adopts the verification methods CV1 and CV2 of the National Construction Code, for buildings on adjoining allotments and on the same allotment respectively.[23] A building must not cause heat fluxes above the prescribed limits at the prescribed distances. It must also be able to withstand the prescribed heat fluxes from a fire in a neighbouring building.[23][5]

In the Netherlands, the building regulations (Bbl) include the preservation of the surroundings, meaning neighbouring plots, as one of their sub-goals. Introduction to Fire Safety Engineering points out that the regulations do not state how large the risk of spread to a neighbouring plot may be.[24] Instead, they set measurable requirements, such as a number of minutes of resistance to fire penetration and fire spread.[24] The IFEG notes that national methods contain assumptions about fire brigade intervention, design radiation levels, flame projection and acceptable damage. For a specific building, it generally prefers a first-principles approach.[19]

Protection measures

The Australian handbook lists the usual measures against fire spread between buildings. These are separation distances, control of the combustibility of external walls, automatic sprinklers and fire-resistant construction.[5] Restricting the size of openings and protecting openings complete the list.[5] Distance works because the view factor, and with it the received radiation, decreases as the distance grows.[16] Where distance is not available, the wall facing the boundary must contain fewer or smaller openings, or the openings must be fire-resisting.[7]

Sprinklers reduce the hazard at its source. C/VM2 uses lower emitted fluxes of 58, 72 and 101 kW/m² for sprinklered firecells, depending on the fire load.[17] When the tables are used, the permitted unprotected area of a sprinklered firecell is twice that of an unsprinklered one.[7] A building more than 1.0 m from the boundary needs no limit on unprotected area at all if it has a reliable sprinkler system and no storage higher than 3.0 m.[17] The sprinkler system must then have a Class A or Class B2 water supply to the New Zealand standard NZS 4541.[17]

Grouped bar chart of emitted radiation: 83, 103 and 144 kW/m² unsprinklered against 58, 72 and 101 kW/m² sprinklered.
Design emitted radiation from unprotected areas in C/VM2 (second edition, 2025) for three fire load energy density bands, unsprinklered and sprinklered. Based on Ministry of Business, Innovation and Employment (2025b), p. 44.

Roofs need attention as well. In an unsprinklered building in New Zealand with an average fire load above 1200 MJ/m², fire spread through a roof within 1.0 m of the boundary must be prevented.[25] This can be done by giving that part of the roof a fire resistance rating, or by calculating the radiation through the roof. A third option is a fire-separating wall along the boundary that rises at least 450 mm above the roof as a parapet, in effect a fire wall.[25] External wall cladding within 1.0 m of the boundary must be non-combustible, of limited combustibility or of material type A.[25]

Limits and uncertainty

The New Zealand commentary states openly that the design radiation values do not guarantee that fire spread is prevented in all situations.[2] Its assumptions include no flame projection from openings, a limited duration of exposure and gas temperatures that follow the standard fire curve. These may not be the most conservative cases.[2] For this reason, the fire service may be needed as a secondary means of preventing spread.[2] The limit of 16 kW/m² at 1 m beyond the boundary is slightly more conservative than the earlier values of 17 to 18 kW/m².[6]

Many factors influence spread in real fires. A study of fires in Tokyo, reported in the SFPE Handbook, considered the distance between buildings and the wind speed. It also included the construction, size and shape of the buildings, the window area and the presence of fences, gardens and trees.[26] The probability of spread between two buildings was described by an exponential function that decreases with the distance between them.[27] Building construction turned out to be the main factor, with wood, mortar and concrete buildings treated as separate groups.[27]

The calculation itself also has limits. View factor calculations for minimum separation distances can be quite complex, especially for irregular facades.[13] C/VM2 therefore requires separate calculations for walls at an angle of more than 10 degrees to the boundary.[7] Firebrands and burning debris fall outside these radiation methods altogether, and their range remains difficult to predict.[13] The approach also depends on the assumed fire. Actual gas temperatures can differ from those in the standard fire test, and the commentary does not allow lower emitted radiation values than those of C/VM2.[6]

See also

Notes

  1. ↑ 1.0 1.1 1.2 1.3 1.4 1.5 1.6 Ministry of Business, Innovation and Employment (2025b), p. 43.
  2. ↑ 2.00 2.01 2.02 2.03 2.04 2.05 2.06 2.07 2.08 2.09 Ministry of Business, Innovation and Employment (2013), p. 131.
  3. ↑ 3.0 3.1 3.2 National Research Council of Canada et al. (2005), p. 1.6-7.
  4. ↑ Australian Building Codes Board (2020b), p. 28.
  5. ↑ 5.0 5.1 5.2 5.3 Australian Building Codes Board (2020b), p. 80.
  6. ↑ 6.0 6.1 6.2 6.3 6.4 6.5 6.6 Ministry of Business, Innovation and Employment (2013), p. 133.
  7. ↑ 7.0 7.1 7.2 7.3 7.4 7.5 7.6 Ministry of Business, Innovation and Employment (2025b), p. 45.
  8. ↑ 8.0 8.1 National Research Council of Canada et al. (2005), p. 1.6-6.
  9. ↑ 9.0 9.1 Lattimer (2002), p. 2-292.
  10. ↑ Ministry of Business, Innovation and Employment (2025b), p. 47.
  11. ↑ Iqbal and Salley (2004), p. F-24.
  12. ↑ Quintiere (2002b), p. 2-255.
  13. ↑ 13.0 13.1 13.2 13.3 National Research Council of Canada et al. (2005), p. 2.6-12.
  14. ↑ 14.0 14.1 National Research Council of Canada et al. (2005), p. 1.6-5.
  15. ↑ 15.0 15.1 National Institute of Standards and Technology (2008), p. xxxvi.
  16. ↑ 16.00 16.01 16.02 16.03 16.04 16.05 16.06 16.07 16.08 16.09 16.10 16.11 16.12 16.13 16.14 16.15 Van Rossum (2027), §5.4.
  17. ↑ 17.0 17.1 17.2 17.3 17.4 17.5 17.6 17.7 17.8 Ministry of Business, Innovation and Employment (2025b), p. 44.
  18. ↑ 18.0 18.1 National Research Council of Canada et al. (2005), p. 3.3-5.
  19. ↑ 19.0 19.1 National Research Council of Canada et al. (2005), p. 2.6-11.
  20. ↑ Ministry of Business, Innovation and Employment (2025b), p. 68.
  21. ↑ Ministry of Business, Innovation and Employment (2025b), p. 67.
  22. ↑ Australian Building Codes Board (2022), p. 14.
  23. ↑ 23.0 23.1 Australian Building Codes Board (2020b), p. 79.
  24. ↑ 24.0 24.1 Van Rossum (2027), §1.2.
  25. ↑ 25.0 25.1 25.2 Ministry of Business, Innovation and Employment (2025b), p. 46.
  26. ↑ Ramachandran (2002c), p. 3-395.
  27. ↑ 27.0 27.1 Ramachandran (2002c), p. 3-396.

References

  • Australian Building Codes Board. (2020b). Handbook: Fire Safety Verification Method (Print version 1.1).
  • Australian Building Codes Board. (2022). Fire Safety Verification Method Standard 2022.
  • 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, Office of Nuclear Reactor Regulation.
  • Lattimer, B. Y. (2002). Heat fluxes from fires to surfaces. 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. 2-269–2-296). National Fire Protection Association; Society of Fire Protection Engineers.
  • Ministry of Business, Innovation and Employment. (2013). Commentary for Building Code clauses C1–C6 and Verification Method C/VM2.
  • Ministry of Business, Innovation and Employment. (2025b). Verification Method C/VM2: Framework for fire safety design (2nd ed.).
  • National Institute of Standards and Technology. (2008). Final report on the collapse of World Trade Center Building 7 (NIST NCSTAR 1A).
  • 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.
  • Quintiere, J. G. (2002b). Surface flame spread. 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. 2-246–2-257). National Fire Protection Association; Society of Fire Protection Engineers.
  • Ramachandran, G. (2002c). Stochastic models of fire growth. 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. 3-381–3-401). National Fire Protection Association; Society of Fire Protection Engineers.
  • Van Rossum, E. (2027). Introduction to Fire Safety Engineering. Fire Safety Academy.
Retrieved from "https://fire-safety.academy/w/index.php?title=Fire_spread_between_buildings&oldid=2570"