1、OR-05-1 5-3 Investigation of Effectiveness of Emergency Ventilation Strategies in the Event of Fires in Road Tunnels Ahmed Kashef, PhD, PEng Member ASHRAE Member ASHRAE Gary Lougheed, PhD Noureddine Benichou, PhD Alexandre Debs ABSTRACT A research project is being conducted at the National Research
2、Council of Canada (NRC) to evaluate the efective- ness 05 current emergency ventilation strategies to control smoke spread in the event of a jre in two road tunnels. The research study includes numerical and experimental phases. e numerical phase uses computational Juid dynamics (CFD) models to stud
3、y smoke ventilation in the tunnels. The experimental phase is used to calibrate and to partially vali- date the chosen CFD models and to provide the necessaly initial and boundary conditions. SOLVENT a CFD model, was used to model ventilation scenarios using existing data. The current paper presents
4、 the eforts to validate the CFD model against on-siteflow andfire test measurements conducted in a 1.8 km road tunnel. The CFD model includes aerodynamically sign$cant physical features of the tunnel and is customized to provide general roughness replicating the actual roughness in the tunnel. INTRO
5、DUCTION Fires in tunnels pose major safety issues and challenges to the designer, especially with the increase in the number of tunnels, their length, and number of people using them. The main fire safety issues include (NFPA 2001; Lacroix 1998): safe evacuation of people inside the tunnel, safe res
6、cue oper- ations, minimal effects on the environment due to the release of combustion gases, and a minimal loss of property. The safety of tunnel users and rescuers is the main objec- tive for the emergency ventilation system (EVS). Life can be threatened in a number of ways: the inhalation of combu
7、stion products, such as carbon monoxide and carbon dioxide, and exposure to high temperatures and heat fluxes. Evacuation can be significantly affected by poor visibility, power failure, blocked exits due to traffic jams or crashed vehicles, or obstruction resulting from a collapse or explosion in t
8、he tunnel. Temperatures up to 1350C and heat fluxes in excess of 300 kW/m2 can be generated within a few minutes of igni- tion in certain types of fires. For safe evacuation, acceptable visibility and air quality must be maintained in the tunnel. From the beginning of a fire, the airflow in a tunnel
9、 is modified and becomes highly unsteady. The modifications are due to the fire itself, the operation of the emergency ventilation system, and the change in the traffic flow in the tunnel. The smoke progress and its degree of stratification depend mainly on the airflow in the tunnel. With no airflow
10、 in the fire zone, the smoke moves symmetrically on both sides of the fire (Heselden 1976). The smoke remains stratified until it cools down due to the combined effects of the convective heat exchange with the tunnel walls and the mixing between the smoke and the fresh air layer. The other parameter
11、s that affect the smoke flow (Heselden 1976) and stratification are fire heat release rate, tunnel slope, and traffic flow. In the event of a fire, the intent of the EVS is to provide tunnel users with a safe egress route that is free of smoke and hot gases. Tunnel operators must implement a plan of
12、 smoke clearing, which consists of selecting a sequence of fan opera- tion with the objective of keeping the road upstream of the fire accident smoke free. This is done by limiting the upstream smoke flow and either venting it using fans or letting it escape through the downstream portal. When the f
13、ire department arrives on the fire scene, the operator must cooperate and modify, as needed, the fan operation in order to facilitate access to the site. Ahmed Kashef, Gary Lougheed, and Noureddine Benichou are with the National Research Council of Canada. Alexandre Debs is with the Ministre des Tra
14、nsports du Qubec, Canada. 1038 02005 ASHRAE. Establishing airflow requirements for roadway tunnels and, consequently, the capacity of the EVS is a challenging task due to the difficulty of controlling many variables. These include changes in traffic patterns and operations during the lifetime of the
15、 facility. Methods of controlling air contami- nants and smoke from a fire in a tunnel using EVS include longitudinal airflow, smoke extraction, and smoke dilution. To evaluate the effectiveness of the emergency ventilation strategies in the event of a fire in the two tunnels in Montreal, Quebec, NR
16、C has undertaken a research project with the Ministry of Transportation of Quebec, Canada. The first two stages of the project have been completed. An extensive literature review on vehicle tunnel ventilation for fire safety has been completed and provided a rational basis for choosing two CFD numer
17、ical models for the initial evaluation: namely, SOLVENT (Innovative Research, Inc.lParsons Brink- erhoff, Inc. 2000) and Fire Dynamic Simulator (FDS) (McGrattan et al. 2000). Based on comparisons with field test data, a model will be selected for use in the remainder of the project. Tunnel Ventilati
18、on System The L.-H.-La Fontaine road tunnel (Figure i), built in 1964, is located in Montreal and travels underwater in a north- south direction. The tunnel is 1.8 km long with three lanes in each direction, inside two concrete tubes. Two ventilation towers are located at the ends of the underwater
19、section. A control and monitoring center for the tunnel is located at the north tower. A central section separates the two tubes. Galler- ies located in this section are used to supply air along the tunnel length via openings distributed along the walls (Figure 2) and these galleries can also be use
20、d as evacuation routes. Doors at various locations along the length of the tunnel provide access to the gallery. There are doors between the galleries providing a route to the other roadway. The wall openings have adjustable dampers to ensure uniform air distri- bution. The side vents are situated i
21、n two rows, upper and lower. The lower and the upper rows are located at heights of 1 .O and 3.9 m above the tunnel floor, respectively, and at inter- vals of approximately 6 m. The two rows of vents are offset by 3 m. The tunnel ventilation is provided by a semi-transverse ventilation system with l
22、ocal extraction points (Figure 1). The ventilation system is composed of eight ceiling exhaust fans (four fans for each roadway) and eight fans that supply air through the side vents, which are uniformly distributed along one wall for each roadway. All fans can operate in reverse mode. Therefore, fr
23、esh air may be supplied at either the ceiling (fans VE151 through VE254) or by fans VA101 through VA204 through the side vents. In the exhaust mode, fans VE15 1 through VE254 can operate at 30 or 60 Hz, and in the supply mode they can only operate at 60 Hz. In the supply mode, fans VA101 through VA2
24、04 can operate at 30,40, or 60 Hz. In the exhaust mode, these fans can only operate at 60 Hz. EXPERIMENTAL WORK Airflow measurements and fire tests were conducted in the north roadway of the L.-H.-La-Fontaine tunnel. The initial airflow measurements (Kashef et al. 2003a) were used to establish the v
25、entilation scenarios for the fire tests (Kashef et al. 2004) and to provide input data for the CFD models. Airflow Measurements On-site airflow measurements were used to determine the capacity of the current EVS of the tunnel and to establish the boundary conditions for the CFD models. Airflow measu
26、re- ments were made throughout the EVS and the tunnel for selected ventilation scenarios. In addition, the range of airflow conditions at the tunnel portals was established. Two fire scenarios were used for the ventilation tests. One fire scenario was located near the mid-tunnel (at a distance in th
27、e range of 555 to 6 15 m from the north portal) and the second close to ventilation fan VE153 (at a distance in the range of 355 to 555 m from the north portal). For these fire locations, South 328 m 350 m 350 m 328 m Slope - 0,25% +0.25% Figure I General layout and ventilation system of the tunnel.
28、 Ventilation and evacuation galleries Upper vents I ppervents South direction North direction 4 12.8 m 6.8 m 12.8 m Figure 2 Typical tunnel cross section. ASHRAE Transactions: Symposia 1039 one emergency ventilation scenario (hereinafter referred to as the “main scenario”) was activated, which used
29、fans VA103 and VA20 1 in the supply mode and fans VE 15 1 and VE 153 in the exhaust mode. After conducting flow measurements of the main scenario, fan VE25 1 was activated in the supply mode, in addition to the already-active fans of the main scenario (hereinafter referred to as the “secondary scena
30、rio”). Airflow temperature and pressure difference measure- ments were made for both the main and secondary scenarios at selected side vents (VA101, VA103, and VA201), ceiling vents (VE151, VE153, VE251, and VE253), and at several tunnel cross sections. Additional measurements were conducted using w
31、eather stations at the two portals, in the unnel, and within the evacuation passage (VA103 gallery). Flow measurements were also conducted at a few locations across the tunnel for the secondary scenario. Air temperatures were used to calculate the air density. The airflow balance in the tunnel was c
32、arefully consid- ered during the tests, especially for the main scenario, as a check of the various airflow measurements. The net airflow (out of the tunnel) was calculated to be about 34.5 m3/s or 43.1 kg/s (assuming an air density of 1.25 kg/m3 at an ambient temperature of SOC). As will be demonst
33、rated in the numerical section, it is believed that this net airflow out of the tunnel can be attributed to an underestimation of airflow at the south portal. Air temperatures and speeds were measured at selected upper and lower side vents in the north roadway. The differ- ential pressure between th
34、e tunnel environment and the gallery was also measured at the lower vents. Measurements were made at side vents located at the beginning, quarter points, and at the end of the ventilation galleries to provide information on airflow in each gallery. The air speed was measured at five points for the l
35、ower vents and at three points for the upper vents. Air velocities were also measured at nine points at the two portals and the middle of the tunnel. In order to determine the bulk flow at a given location, the cross section was divided into three, five, or nine subareas associated with the measurem
36、ent points. The velocity of air at each point was considered to represent the entire subarea. The bulk flow was then computed by multiplying the air velocity by its corresponding subarea and then adding the results for all subareas. A positive air velocity within the tunnel represents airflow from t
37、he north portal toward the south portals. The error in the values of bulk flow rates calculated using this method is estimated to be in the range of2% to 4% as there was an allowance for the effect of the turbulent boundary layer flow. A summary of airflows in the tunnel for both the main and second
38、ary scenarios is shown in Figures 3 and 4, respec- tively. The pressure inside the ventilation gallery must be larger than that in the tunnel roadway to ensure that the airflow is directed from the gallery into the tunnel. This would also ensure a smoke-free escape route in the gallery for tunnel us
39、ers. In general, the main flow direction was from the venti- lation gallery into the tunnel area. This was confirmed by I 211 South Portal VA-I03 VA-201 I II Ndhi%wtd sesign 230 Figure 3 Airflow measurements for the main scenario. VE-251 VE-253 rn 137 NIA MeaSuf9d 141 * NIA I II I I *Measured using
40、lhe vane onemomeler South Portal Volumebic flow volues are in m/s Figure 4 Airflow measurements for the secondary scenario. releasing cold smoke bombs at different locations in the tunnel, representing smoke and fire effluent sources. No smoke was observed within the two ventilation galleries and th
41、e results indicated that the airflow was directed into the tunnel. Ventilation fan VA101 was not activated as part of the selected ventilation scenario. However, low positive airflow velocities were measured at the side vents in this section of the supply system. The airflow conditions were measured
42、 within the ventilation gallery of fan VA103. The air speed was 14.5 ds, indicating a dynamic pressure of approximately 13 1 Pa. During the cold smoke tests, it was observed that smoke was efficiently cleared when the smoke source was placed close to exhaust fans VE151 and VE153. However, when the s
43、ource was placed near the middle section of the tunnel, it took longer for the smoke to clear as a result of low airflow veloc- ities in this section of the tunnel. This suggested that other ventilation scenarios should be explored for a fire located near the middle of the tunnel. 1040 ASHRAE Transa
44、ctions: Symposia Figure 5 Fire tests. By activating fan VE25 1 in the supply mode, higher air velocities were recorded near the middle section of the tunnel (Figure 4). This, in turn, improved the smoke clearance. However, in examining the airflow distribution, only a small portion of the air suppli
45、ed by VE25 1 was directed toward the middle section of the tunnel with the remaining portion of supplied air directed toward the south portal. This increased the airflow exiting the portal from 72 to 189 m3/s. Also shown in Figures 3 and 4, the airflow introduced at the side vents by fan VA103, toge
46、ther with the air pulled through the north portal, was comparable to the airflow exhausted by fans VE 1 5 1 and VE 153. As a result, minimum airflow was drawn from the middle of the tunnel, causing low airflow velocities at mid-tunnel and beyond. Even with the activation of fan VE25 1 in supply mode
47、 (secondary scenario), the airflow was mainly directed to the south portal, resulting in little improvement in airflow velocities at the middle of the tunnel. It is believed that most of the flow exiting the south portal was fresh air supplied by fans VA20 1 and VE25 1. Fire Tests Fire tests were co
48、nducted using a propane burner system developed by NRC. Two fire tests were conducted in the north roadway of the tunnel (Figure 5): one in the middle of the tunnel and one close to the exhaust fans at the north end of the tunnel. Two different EVS scenarios were activated for the two tests. For Tes
49、t 1, the EVS scenario used the two fans VE151 and VE153 in the exhaust mode and the two fans VA103 and VA201 in the supply mode. In Test 2, a similar EVS scenario was activated except for fan VA103, which was used in the exhaust mode. During the planning phase, the project team investigated several issues to develop the fire source for the in-situ fire tests. These issues included: fire source alternatives, fire size, test duration, tunnel safety, and test logistics. In order to simulate a fire, the heat source must produce a buoyant airflow. Therefore, the heat source mus
