Reduced-scale experimental and numerical study of fire in a hybrid ventilation system in a large underground subway depot with superstructures under fire scenario

https://doi.org/10.1016/j.tust.2019.02.006Get rights and content

Abstract

To reduce energy consumption and improve the safety and reliability of ventilation systems in underground subway depots, a hybrid ventilation (HV) system composed of natural ventilation, mechanical fans, and flow deflectors was proposed and tested. A reduced-scale (1:50) experiment and a full-scale numerical simulation were conducted, wherein four heat release rates (HRRs; 281 W, 380 W, 531 W, and 866 W) of the fire source and five ventilation velocities (0.7 m/s, 1.0 m/s, 1.4 m/s, 1.9 m/s, and 2.4 m/s) were tested. The temperature distributions under the ceiling were measured. Smoke movement and smoke layer stability were visualized using a laser sheet. The smoke layer height, gas flux of shafts, and smoke movement route were recorded from the simulation. Under HV, the ceiling temperature decreased significantly with increasing ventilation velocity; however, changes in temperature were different at different locations. With appropriate ventilation velocity (1.4 m/s), HV effectively controlled the smoke temperature of the bottom layer and ensured the stability of the smoke layer in the interlayer. However, the stability of the smoke layer was disrupted at higher ventilation velocities, which caused the smoke to sink, whereas a lower velocity could not slow the rise in temperature. The relationship among ventilation velocity, HRR, ceiling temperature, and smoke layer stability was analysed. A new criterion, N (N = 0.62), was proposed to determine the critical ventilation velocity associated with a lower ceiling temperature and improved smoke layer stability.

Introduction

As urbanization continues in China, land resources are becoming increasingly scarce, and an increasing number of urban buildings are underground. Many underground depots with residential and business buildings above them have been constructed in recent years (Fig. 1) such as the Beijing subway line 7 depot, the Nanjing Hexi tram depot, the Shenzhen subway line 11 depot, and many others in Hong Kong, Hangzhou, and Suzhou (Xu et al., 2014). The buildings located above an underground depot are called superstructures. Subway and tram depots are generally substantially larger than ordinary underground stations, and the areas comprising these structures are necessarily interconnected and cannot be isolated. Therefore, in the event of a fire, smoke can spread rapidly. Ventilation is considered the most effective way to control the spread of smoke. To date, mechanical ventilation (MV) systems have not been specified as a necessary installation for such buildings in the relevant design specifications (GB 50016, 2014, GB 50157, 2003). Therefore, a few depots use natural ventilation (NV) systems, whereas others use MV systems. However, our previous investigation revealed that the performance of traditional NV systems is unstable and inefficient in the event of a fire in such buildings (Xu et al., 2014). When the heat release rate (HRR) of the fire source is not sufficiently high, buoyancy alone cannot ensure that the smoke moves outside the building. Moreover, the associated large area means that an MV system requires a substantial initial investment in addition to operational and maintenance costs. Therefore, a reliable and economical ventilation system must be developed.

In recent years, many studies on smoke control have been conducted in underground buildings such as garages and subway stations. Chen et al. (2003) investigated the effectiveness of the smoke control scheme in a typical subway station, and the results indicated that the stack effect plays a deterministic role in smoke control when a fire occurs near the stairs. When a fire occurs in other places, the mechanical control schemes of the subway station are effective. An experimental study of the stack effect in tall vertical shafts using scale modelling was reported by Chow and Zhao (2011), and their results confirmed that hydrostatic equations are acceptable for studying stack pressure in tall vertical shafts. Zhang et al. (2007) conducted a numerical simulation to determine the speed at which smoke would move in an underground garage under different types of ventilation. Ji et al. (2011) conducted a reduced-scale experiment and made numerical predictions to calculate the maximum temperature near the ceiling in an underground subway station. In addition, smoke stratification in a horizontal channel was investigated by Tang et al. (2017), and buoyancy and inertia force were correlated through the Froude number and the Richardson number. Then, the range in which smoke can remain stable was obtained. Gao et al., 2012a, Qin et al., 2009 studied the spread of smoke in different fire scenarios in a building with a large space. Although the structures considered in these studies were not particularly similar to subway depots, the results revealed the basic rules of smoke movement in underground buildings. The performance of hybrid ventilation (HV) in controlling fire-induced smoke in subway stations and tunnels has been studied by some scholars Luo et al., 2014, Gao et al., 2012b, Tanaka et al., 2015, and it was found that HV can effectively confine smoke. Most of the previous studies have focused on traditional buildings such as subway stations, tunnels, and garages. Therefore, in the present study, we investigated whether an HV system can be applied to control fire-induced smoke in an underground depot.

To answer this question, a new HV structure was proposed by combining the structural characteristics of an underground depot with a superstructure. To maintain building fire safety and reduce investment in the construction of a ventilation system, the aforementioned HV structure, which is composed of shafts and mechanical devices, was introduced into this type of building. The shaft in an HV system is a structure typically associated with an NV system; it accelerates the flow of hot gas due to the temperature difference between the upper and lower openings (Klote, 1991). Because of this function, shafts have played crucial roles of routine ventilation and smoke dissipation in ventilation systems Vauquelin and Telle, 2005, Sanada et al., 2015). However, a shaft’s efficiency is affected by its geometry and the HRR of the fire source (Yuan et al., 2015). Therefore, many studies have been conducted to improve shaft efficiency, and criteria and models have been proposed such as critical shaft height (Ji et al., 2012), the stability criterion for convection in a shaft (Kazakov et al., 2015), the vertical distribution model of temperature (Chen et al., 2016) and the ratio between building the height and mass flow rate of the shaft (Giachetti et al., 2016). These results are critical references for our study and are applied to the installation of shafts without mechanical fans. Shafts with mechanical fans are also widely used for smoke extraction in the event of a tunnel fire. Jiang et al. (2016) conducted an experiment to study the air entrainment coefficient and the increase in smoke spread velocity induced by exhaust fans installed in shafts. The critical volume rate of smoke exhaust that leads to plug-holing in a tunnel was determined, and this measure served as a reference for setting the wind velocity in HV experiments. Moreover, the smoke accelerated by a mechanical fan in a shaft is similar to a hot plume. When the smoke reaches the ceiling, a ceiling jet that carries combustion products will also be produced. It is important to understand the characteristics of a ceiling jet’s flow, which runs along the ceiling of the interlayer. A systematic method for studying the characteristics of ceiling jets was reported by Oka et al. (2016).

Although nearly every aspect of HV, such as shafts and mechanical fans, has been studied under different scenarios, limited research has evaluated the behaviour of HV in the context of an underground depot. Moreover, certain properties of buildings considered in the previous studies are different from those of an underground depot with a superstructure. Indeed, underground subway depots have several distinct characteristics, including a large space and dense connectivity. Therefore, in the present work, a systematic investigation was conducted to discover the dynamics of smoke movement in an underground depot with HV and to evaluate the performance of HV under different fire scenarios. The temperature distribution under a ceiling with multiple shafts was also measured. According to the principles of experimental simulation and recommended model parameters (Vauquelin, 2008, National Fire Protection Association, 2009), a reduced-scale experimental model was established. A series of experiments was then conducted to study the ceiling temperature distribution and smoke layer stability. A numerical simulation was performed to study smoke velocity, outdoor air and smoke movement route, and shaft flux. The criterion N was derived using the obtained experimental and simulation results.

Section snippets

Experiment and simulation model

As shown in Fig. 2, the depot comprised two levels: the bottom layer (one level below ground) and the interlayer (between the bottom layer and the superstructure). The superstructure, including residential and business buildings, was above the interlayer. A total of 12 shafts were evenly arranged in the interlayer, and each of the two shafts at the centre of the interlayer was fixed with a mechanical device composed of an exhaust fan and a flow deflector (Fig. 3). The bottom layer included the

Experimental results and discussion

A comparison of the smoke temperatures between different ventilation systems and HRRs is shown in Fig. 9. This study featured 24 cases. To concisely explain each case, 3D curves were used, and the 2D data curves are shown in the supplements. To show the data in each subfigure as clearly as possible, each subfigure was created from a slightly different perspective. The experimental results for HV under the five wind velocities are explained.

As shown in Fig. 9a, the temperature increased as the

Conclusion

  • (1)

    The experimental results show that the temperature of the bottom layer in HV was much lower than that in NV at all HRR values from 280 to 866 W, indicating that an HV system controls the temperature more effectively than does an NV system.

  • (2)

    Smoke diffusion along the ceiling in the bottom layer could have been suppressed by the flow of outdoor air into the exhaust fan, and the smoke did not accumulate in the interlayer.

  • (3)

    When the wind velocity was less than 1.0 m/s, not only did the temperature of

Acknowledgements

This work is supported by National Natural Science Foundation of China [grant number 51476075, 51508267] and A Project Funded by the Priority Academic Program Development of Jiangsu Higher Education Institutions.

References (34)

Cited by (20)

View all citing articles on Scopus
View full text