FPEeXTRAIssue47

A Method to Estimate the Effect of Wind on Natural Ventilation in Large Spaces

By: Nils Johansson

The influence of wind is only occasionally accounted for in fire safety engineering, still, its effect on, for example, the performance of a smoke management system can be large. There are simple and straightforward calculation methods available for the design of smoke control in large spaces, like the iterative method by Yamana and Tanaka [1]. However, such methods do not take external conditions like the wind into account. In a recent paper [2], presented at the 15th International Conference and Exhibition on Fire Science and Engineering, a modified version of the Yamana and Tanaka method, which accounts for wind effects, was presented and evaluated.

Theory

In the original method by Yamana and Tanaka it is assumed that that smoke is vented through ceiling vents and that the smoke layer stabilizes at a certain height, z. At this point, the plume mass flow, , equals the total mass flow that leaves the hot layer through the ceiling vents, (see Figure 1). The plume mass flow is calculated with a suitable plume model. Vigne et al [3] studied several different plume models and found that the McCaffrey correlation gave results closest to experimental data after a hot gas layer stabilized. It not considered possible to conclude that the McCaffrey correlation always will be preferred; however, the scenario studied by Vigne et al resembles the situation under study and the McCaffrey correlation was therefore used in this study.

Figure 1: Principles of the Yamana-Tanaka method.

In the modified method, the effect caused by wind on natural smoke ventilation is approximated with the geometric configuration of the building and position of the make-up air and smoke vent regarding the wind direction. For a simple scenario (rectangular building free of local obstructions), the excess pressure caused by wind on the exterior can be calculated with the following equation.

Where v is the wind speed and μ is a shape factor that depends on the shape of the building. The size of the shape factor will depend on the geometry of the building. The shape factor will be positive on the surfaces of the building that is facing the wind (windward) and negative on the leeward sides where the wind will cause a suction effect. The pressure difference caused by the wind on the roof vents () and the inlet vents () can be added to the pressure difference in the “zero wind” case as presented in equations 2 and 3. The full method is presented in the original paper [2].

In a building with horizontal openings for natural ventilation on the roof, negative wind pressure on the roof will be beneficial, since it will increase the pressure difference over the opening. On the contrary, wind can have negative effects on smoke extraction if it reduces the pressure difference over the make-up air openings.

Results from the presented method were compared to results from the Fire Dynamics Simulator (FDS). A limited number of cases were studied with FDS (see Table 1); however, the cases were designed to cover different types of ventilation (see Table 2) and wind conditions. A windless scenario was also studied for each case, as well as two different wind directions, i.e. wind on the windward or leeward side of the make-up air opening. This means that a total of 36 scenarios were studied.

Results

The calculated vent mass flows are compared in Figures 2 and 3. “Wind-ward” and “Lee-ward” in the figures refers to the position of the make-up air in regard to the wind direction.

In the windless situation there is a consistent deviation in the results between the studied methods, i.e. FDS gives roughly 15% higher vent mass flow (see Figure 2, left). This should not be considered as a general difference between the two methods since it is related to the studied scenario. However, the fact that the shift is consistent implies that the relevant phenomena are captured in the analytical method for windless scenarios. This also seems to be the case for low wind speeds where FDS consistently gives a 10-15% higher vent mass flow (see Figure 2, right).

As the wind speed increases, the consistency between the models decreases (see Figure 3). A contributor to this is the fact that the influence on the fire, the plume and the conditions in the room increases with increased wind speed, an effect that is entirely ignored in the analytical method.

The analytical method demonstrated that all the wind conditions studied were beneficial in regard to the mass flow through the openings for natural ventilation. This is not seen in the FDS simulations, where a decreased vent mass flow was seen as the wind speed increased.

The influence of wind on the exterior and interior of a building is complex and the analytical method allows for swift calculations of how to smoke filling can be affected by different wind conditions. However, the simplifications included in the analytical method might be too large for the method to be useful. Even so, only a limited number of scenarios were studied in the paper, and further evaluation of the method with both CFD modeling and experimental data is needed to frame its possible area of use.

Nils Johansson is with Lund University.

References

  1. Yamana, T. and Tanaka, T., Smoke Control in Large Scale Spaces, Fire Science and Technology 5:31-40, 1985. https://doi.org/10.3210/fst.5.31
  2. Johansson, N. & Karlsson, B., An Analysis of a Method to Estimate the Effect of Wind on Natural Ventilation in Large Spaces, In 15th International Conference and Exhibition on Fire Science and Engineering, London, UK, 2019.
  3. Vigne, G., Gutierrez-Montes, C., Cantizano, A., Węgrzyński, W. and Rein, G., Review, and Validation of the Current Smoke Plume Entrainment Models for Large-Volume Buildings, Fire Technology, 55:789-816, 2019. https://doi.org/10.1007/s10694-018-0801-4