Influence of the Building on a Design Fire
By Jonathan Wahlqvist and Patrick van Hees
Division of Fire Safety Engineering
Lund University
Design fires are often used to evaluate performance-based designs by fire safety engineers all over the world and can be an invaluable tool if used properly. Recently, there has been a tendency for recommendations on design fires to be included in the building codes to assist the designer and narrow down the possible variation between different designers. However, one potential issue with this is that the same design fire is recommended in similar building types despite the fact that some building characteristics that can affect the fire development, such as building material, can differ greatly. This paper gives a summary of a numerical investigation by Wahlqvist and van Hees1 of how several key characteristics of a building influence the design fire.
The simple design fire concept has been used extensively to evaluate performance-based designs all over the world. However, the design fire is a rough simplification of the real world, and using it in applications outside the boundaries of its original intent might result in erroneous conclusions in regard to fire safety. However, a problem that can arise from directly using proposed growth rates is that the characteristics of the fire compartment are never accounted for. For example, will a heavily insulated building behave the same as a steel sheet building or does the radiative feedback increase the fire growth and maximum heat release rate? Does a smaller room behave different from a big room? How much does the amount of openings (both normal openings and those caused by evacuating people, as in doors being opened) to the fire room affect the development of a fire?
Scenarios
To investigate the influence of different building characteristics on design fires, a simple room was created in the Fire Dynamics Simulator (FDS). Simulations were performed with different building materials, floor areas, room heights and openings. However, a room with the dimensions 10x10x3 meters (see Figure 1) was used as reference. The 10x10x3-meter compartment was selected as the “default” room to represent a reasonable “normal” case, but it was also adapted to be able to see a clear distinction between different building materials.
Figure 1: The 10x10x3-meter room with one door opening and the fire source placed in the
middle of the room
Numerical Model
To be able to predict the changes in growth rate and maximum heat release rate the radiative feedback to the fuel and oxygen depletion is important to model. A simple model, implemented in FDS, used to dynamically change the heat release rate has been presented and validated by Wahlqvist and van Hees2. In short, the model consists of two distinct parts: lowered mass loss rate due to lowered oxygen levels close to the fire source and increased mass loss rate due to radiation from external sources (e.g., walls and smoke layer).
The oxygen volume fraction at the flame base is used to describe the change of radiative feedback to the fuel caused by cooling of the flame, extension of the flame or detachment of the flame from the pool surface. The reduction in radiative heat flux feedback in turn results in lowered mass loss rate. Because simulating the radiation feedback from the flame can be very challenging, using the oxygen fraction at the flame base can potentially represent this behavior in a simplified model.
In many general cases, the temperature of the fire room walls and smoke layer do not heat up enough to re-radiate any significant amount of energy to the fire source; this could either be due to well-ventilated conditions with a lot of air exchange or large heat losses through the compartment boundaries. But in some cases, such as sealed compartments or well-insulated compartments as used in this study, this re-radiation can amount to a significant portion of the total mass loss rate and in some cases even be the dominating contributor.
Results
Influence of building material and growth rate
It was observed that there were minor changes in the growth rate due to building material (concrete 1–7%, drywall 8–35% and steel 0–20%) with the exception of the insulated walls, where the difference in growth rate is at most almost 300% and the least about 115%. The actual growth rate was 0.012 kW/s2 when using the insulating walls and specifying a growth rate of 0.003 kW/s2, the same as the next standard classification “medium.” The relative increase is not as great (114–213%) using the higher growth rates, but it could still be considered significant when used to assess the fire safety in a building.
It can also be noted that the maximum heat release increased with increased growth rate for all building materials. However, the peak heat release rate duration is rather short because the oxygen depletion starts to affect the mass loss rate rather quickly. It is evident that the growth rate was hampered by the lack of oxygen being supplied to the compartment (see Figure 2).
The fact that the maximum heat release is increased with increased growth rate is simply due to the fact that a faster growth rate reaches a higher heat release rate before the effects of oxygen depletion sets in.
Figure 2: Heat release rate as a function of time using different building materials,
one door opening and a growth rate of 0.003 kW/s2
Influence of door openings
The growth rate was larger in all cases when four doors were present, compared to only one open door. For example, the growth rate when using the insulating walls and specifying a growth rate of 0.012 kW/s2 now exceeds the next standard classification (fast, 0.047 kW/s2) with a calculated growth rate of 0.057 kW/s2 (see Figure 3). In fact, all specified growth rates are higher or close to the next standard classification (when available). This would probably mean that the time available to perform safe egress from a building would be overestimated.
Figure 3: Heat release rate as a function of time using different building materials, four door
openings and a growth rate of 0.003 kW/s2
Influence of the room floor area
It was observed that the room floor area can have a significant influence on the growth rate and maximum heat release rate. In almost all cases the main driving force is the radiative feedback because the oxygen volume fraction close to the fire source reduces to a value close to or below the limit for extinction.
The external radiation heat flux does get significantly larger in the cases with four door openings due to the same reasons given before; because more air is supplied, more combustion can occur inside the compartment and closer to the fire source, and this increases the temperature of the walls and hot gasses. Hence, the radiative feedback intensifies. Another interesting observation is on the maximum heat release; when only one door opening is present the highest recorded heat release rate is in the largest floor area but only for a very brief time. When four door openings are present the smallest floor area reaches the highest heat release rate and over a much longer period of time (see Figure 4).
Figure 4: Heat release rate as a function of time using different room floor areas, four door
openings and a growth rate of 0.003 kW/s2
Influence of the ceiling height
It was observed that the room height has a significant effect on the growth rate and maximum heat release. Having only one door opening, the growth rate drastically decreases as the ceiling height increases. This is due to the decrease of external radiative heat flux feedback with increasing ceiling height in combination with the fact that some degree of oxygen depletion occurs regardless of ceiling height. When four doors are present the effect is very similar although to a larger degree. However, the effect of the room height is not as significant if the room floor area is smaller because the walls seem to contribute to a larger degree. A combination of a larger floor and high ceiling height actually renders very little or almost no radiative feedback which might be “good news” for scenarios in which performance-based design often is applied (big open spaces). It should, however, be remembered that nearby objects or walls still can provide radiative feedback.
Conclusion
It has been demonstrated that the building material used in a fire compartment might influence the growth rate of a design fire in a significant way; insulating wall/ceiling materials will probably increase the growth rate, which in turn will mean that the time to critical conditions will be shorter compared to the expected result. However, if the building material used is thermally thin or has a large heat-storing capacity, the influence is small. The maximum heat release rate and transient behavior of the fire was shown to be very dependent on ventilation factor in combination with the building material. With only one door opening present the initial fire growth was rapid due to radiative feedback but was then hampered by the oxygen depletion, which caused the heat release rate to decrease significantly during the course of the simulation duration.
Further, it was shown that the room floor area might have a significant effect on the growth rate and maximum heat release rate. With one door opening a larger room would have a higher initial peak heat release rate than the smaller rooms, but they would all get oxygen depleted soon thereafter and behave similarly for the rest of the simulation. If there were four door openings present the radiative feedback would overpower the oxygen depletion and the smaller the room the higher the maximum heat release rate and growth rate.
It was also shown that ceiling height does have a significant effect on the growth rate and maximum heat release when coupled with four door openings. However, the effect of the room height was not as significant if the room floor area was smaller because the walls seem to contribute to a larger degree in that case.
In conclusion, it is recommended to investigate the applicability of design fires in compartments with highly or moderately insulating building materials, as the environmental feedback will probably increase the growth rate significantly, which in the end could affect the possibilities of obtaining satisfactory egress safety.
This paper is just a summary of research done by Wahlqvist and van Hees. If you want to find out more, read the full paper1 published as open access in the journal Case Studies in Fire Safety.
References
[1] Wahlqvist, J. and van Hees, P. “Influence of the built environment on design fires.” Case Studies in Fire Safety (5), pp. 20–33, 2016.
[2] Wahlqvist, J. and van Hees, P. “Implementation and validation of an environmental feedback pool fire model based on oxygen depletion and radiative feedback in FDS.” Fire Safety Journal (85), pp. 35–49, 2016.