BUILD SIMUL (2013) 6: 173 – 182 DOI 10.1007/s12273-013-0101-3
A network-based smoke control program with consideration of energy transfer in ultra-high-rise buildings, CAU_ESCAP Research Article
Sungryong Bae1, Gwon Hyun Ko2, Chang Wook Lee2, Hong Sun Ryou1 () 1. Department of Mechanical Engineering, Chung-Ang University, Republic of Korea 2. Department of Architecture and Fire Administration, Dongyang University, Republic of Korea
Abstract
Keywords
Ultra-high-rise buildings allow for the efficient use of land, but they are vulnerable to disasters such as fires. Therefore, the development of network models for analyzing the characteristics of smoke movement in ultra-high-rise buildings is necessary for cost-effective design of smoke control systems and operation decisions. A new network-based smoke control program, CAU_ESCAP, is developed in this study, which is a program that can consider the energy transfer. CAU_ESCAP is validated with existing programs, ASCOS and COSMO, by analyzing the smoke movement. After that, fire in an ultra-high-rise building of 55 stories is applied with CAU_ESCAP for analyzing the smoke movement and the mass flow rate of the smoke control system due to the variation of heat release rate and door conditions of the fire floor. The pressure difference between the fire room and the protecting area does not vary in the closed-door case in the fire room, but vary significantly in the opened-door case. Therefore, the smoke from fire would be spread to other spaces if there is no instantaneous increase in the mass flow rate of pressurization when the door is opened by occupants for evacuation.
ultra-high-rise building,
1
Introduction
E-mail:
[email protected]
network-based program, energy transfer, CAU_ESCAP
Article History Received: 14 March 2012 Revised: 26 November 2012 Accepted: 27 November 2012 © Tsinghua University Press and Springer-Verlag Berlin Heidelberg 2013
California fire, One Meridian Plaza fire, Cook County Administration fire, and the Parue Central Building fire. These fires resulted in high casualties because the smoke had spread rapidly throughout the entire building via vertical shafts such as elevators, stairways, and the HVAC (heating, ventilation, and air conditioning) systems (Yu et al. 2007). Recently, a huge fire occurred in the Woo Sin Golden Suite, which is an apartment building in Busan, R.O. Korea. The fire propagated through the inflammable outer wall, but the smoke from the fire spread throughout the building via the ducts of the HVAC system, and poisonous gas obstructed efficient evacuation (Choi 2010). It is clear from these cases that effective smoke control system designs are needed for reducing the number of casualties resulting from poisonous gas in ultra-high-rise building fires. The smoke movement in building fire situations is sensitively influenced by the structure, environmental conditions, pressurizing methods, and the fire conditions. As such, the design of smoke control systems for controlling
Indoor/Outdoor Airflow and Air Quality
Recently, increases in population density due to metropolisbased industrialization have led to high density and the integration of living environments. The construction of various ultra-high-rise buildings has been increasing to accommodate the growing number of people in cosmopolitan cities, as well as to make effective use of land (Ministry of Land 2009). The ultra-high-rise buildings allow for efficient use of the land, but they are vulnerable to disasters such as fires. Particularly, should the exterior of an ultra-high-rise building be broken in a fire situation, the broken area would become a supply route of oxygen, and the stack effect in shafts then becomes stronger because of the characteristics of the sealed exterior. Also, the velocity of the smoke spread would increase through vertical shafts such as elevators and stairwells, and this phenomenon would cause heavy casualties (Park 2009). Representative examples of fires in high-rise buildings include the MGM Grand Hotel fire, First Interstate Bank of
smoke control system,
174
and discharging the smoke is a difficult work. Therefore, a great deal of research has been carried out to understand the characteristics of smoke movement and to design smoke control systems. CFD methods that have been validated with experimental data, such as FDS and Fluent, are primarily used for assessing the fire safety of buildings, because of the high accuracy that can be achieved in accounting for the characteristics of heat and smoke movement. However, these methods have high computational cost because of the large number of small computational grids. Therefore, the development of network-based models that can be used for analyzing the characteristics of smoke movement in ultra-high-rise buildings is necessary for the cost-effective design of smoke control systems and operation decisions. Klote at al. (Klote and Fothergill 1983; Klote and Bodart 1985; Klote 1993, 1995) developed a network-based program, ASCOS, which assumed that each compartment is represented as a network or node, and that the airflows would be developed by pressure differences between each node. ASCOS was validated with experimental results, and widely used for designing smoke control systems. Walton (1989) developed the flow calculation algorithm AIRNET to increase the computational speed. After that, Waton (1994) developed the CONTAM 93 program, with an improved version of the AIRNET flow routine and easy input. Recently, Wang et al. (2011) developed a new version of CONTAM, CONTAM 3.0, which applied a CFD model, CFD0, for calculating the outdoor flow field and the mass transfer in the building. Then, using CFD0, the results for the flows internal and external to the space of the building and the distributions of pressure in the building were improved. The established programs, like ASCOS, the CONTAM series, and others, calculate the airflow from the pressure and temperature data of each compartment. However, the temperature profiles of a building must be assumed, or they must be determined in separate programs, and the energy transfer from the fire is neglected. These estimated temperature distribution results in decreased numerical accuracy, and therefore, to supplement this shortcoming, Black (2009, 2010) developed the COSMO program, which combines the advantages of a network model and a differential approach. COSMO has an advantage for estimating suitable results with fire phenomena, because it estimates the smoke movement with temperature, pressure, and density, through the conservation of mass, momentum, and energy, and the equation of state. However, COSMO cannot simulate the fire exactly, because there is no function for exactly simulating the fire in the program, and the airflows between each node are then calculated with the assumption that the fire is expressed by the pressure and the temperature. It also has a disadvantage regarding computational time and memory,
Bae et al. / Building Simulation / Vol. 6, No. 2
because an iterative method is used for solving 13 equations and 13 variables of convective and radiative heat transfer (Black 2010). Therefore, a new network-based model for simulating the smoke movements in ultra-high-rise building fires has to be developed for improving the computational cost and the accuracy of simulation results. 2
Development of CAU_ESCAP
2.1 Theoretical basis Network models are more appropriate for simulating conditions in a large structure, such as ultra-high-rise buildings, because they have high computational efficiency. Each building compartment is represented as a network or node, and each node has a uniform set of properties, such as pressure, temperature, and density (Klote and Fothergill 1983). The mass flow rate is then calculated using the properties of each node, and the entire node must satisfy the conservations of mass and energy. Then, conservation of mass in the unsteady state is presented as (White 2008) Nc
No
i =1
i =1
å m ( i , j ) + å m o( i ,k ) + m f ( i ) = 0
(1)
where, m ( i , j ) is the mass flow rate between node i and node j, m o( i , k ) is the mass flow rate between node i and node k (outdoor node), and N c and N o represent the number of flow passages of node i to indoor node j and to outdoor node k, respectively. The source of the mass flow rate from the fire can be calculated from the fire’s heat release rate (HRR). The relation between the heat release rate of the fire, Q , and the mass loss rate of fuel is represented as (Quintiere 1998) m Fuel =
Q = m f ( i ) DH c
(2)
where, m Fuel is the mass loss rate of fuel, and DH c is the heat of fuel combustion. We assumed that the mass loss rate of fuel is converted into the mass flow rate of the source, and m f( i ) is the mass flow rate of the source from the fire in Eqs. (1) and (2). The convective heat transfer mainly affects the heat transfer between each node, so it is assumed that the convective heat transfer process between each node is under quasi-equilibrium. Therefore, the heat transfer is equal to the change in enthalpy, and enthalpy conservation in the process of quasi-equilibrium is (Sonntag et al. 2005) Nc
å m
No
h + å m o( i , k )hio + m f ( i )hf ( i ) = 0
(i, j ) i
i =1
(3)
i =1
where, m ( i , j ) and m o( i , k ) represent the mass flow rate
175
Bae et al. / Building Simulation / Vol. 6, No. 2
between node i and j or k node (outdoor node), and calculated values from conservation of mass; hi and hio are the enthalpy of node i or the enthalpy of node j or k and hf ( i ) is the enthalpy from the fire. Because the radiative and conductive heat transfer affect the heat increase in the fire node, those heat transfer is included in the heat flux from fire. Then, the enthalpy of both source and each node can be calculated using the temperature of the flame and the node. The relation between the enthalpy and the temperature is represented as (Sonntag et al. 2005) hf = 0.00009630Tf2 + 0.94032579Tf + 9.13398946
(4)
where, hf is the enthalpy of the source, and Tf is the temperature of the flame. Figure 1 is a schematic of the mass and enthalpy flow calculation at each node. As shown in the figure, one of the compartments is regarded as one node, and the sign of the inflow is considered to be (+), while the outflow is considered to be (–).
Fig. 1 Schematic of the mass and enthalpy flows Fig. 2 Flow chart of the method used by CAU_ESCAP
2.2
Numerical algorithm of CAU_ESCAP
Figure 2 shows a flow chart of the method used by CAU_ESCAP. As shown in the figure, the pressure and mass flow rate of each node are calculated first. After that, the energy transfer from fire is calculated, and the pressure and temperature are then coupled by using the density of each node. Part I of Fig. 2 is a loop for calculating the mass flow rate and pressure distribution of the building. These calculations are done separately for the building compartment and shaft compartment. In the calculation of the building compartment, the orifice equation is selected for calculating the mass flow rate of air, Q, through the leakage area between neighboring nodes (White 2008): Q = CA
2( Pi - Pj ) ρ(1 - β4 )
(5)
where, A is the leakage area between neighboring nodes, Pi and Pj are the pressures of each node, β is assumed to
be zero because the leakage area is negligibly small compared to the wall, and C is the flow coefficient and generally taken to be in the range of 0.6 to 0.65 (White 2008). In the calculation of the shaft compartment, the shaft is considered to be one node, and the pressure distribution by height is calculated by Pj = Pj-1 - Pz - Pf
(6)
where, Pj is the pressure at the j-th floor, Pz is the hydro static pressure difference, and Pf is the pressure loss due to friction. Each pressure represented as Pz =
gP ( z j - z j-1 ) RT m u
2
(C )
Pf = S
(7)
(8)
s
where, z j is the height of node j, g is the acceleration of
176
Bae et al. / Building Simulation / Vol. 6, No. 2
gravity, R is the universal gas constant, P and T are the averaged pressure and temperature, S is the direction of the mass flow, m u is the mass flow rate of the upward directed flows in the shaft, and Cs is the flow coefficient in the shaft (Klote 1991). The influx and outflow of the shaft compartment are then calculated by Eq. (5) using the node neighboring the shaft compartment. The calculation of the mass flow rate is iteratively solved until the sum of all mass flow rates at all of the nodes becomes lower than the convergence limit of 0.00002 kg/s. Part II of Fig. 2 is a loop for calculating the energy transfer. In contrast to part I, the calculation of energy is not separated between the building and shaft compartments, and the shaft compartment is divided into each floor. Equation (3) is iteratively solved until the sum of fluxes becomes lower than the convergence limit of 10 J/s. Part III of Fig. 2 is a loop for calculating the density of each node from the energy transfer calculation using distributions of pressure and temperature and checking the convergence of all calculations. For calculating the density of each node, we assumed that the pressure of every node depends on the temperature only, since the variation in pressure has a negligible influence on the density change compared with the variation in temperature. In such a case, the relation between temperature and density is represented (Schwarz and Janicka 2009): ρ = 360.77819T -1.00336
(9)
where T is the temperature of each node in Kelvin. The relative error is then calculated to check the convergence, as presented in (Chapra and Canale 2005): εa =
ρn+1 - ρn ρn+1
(10)
where, ρn is the density before the calculation of energy transfer, and ρn+1 is the density after calculation. In CAU_ESCAP, the convergence criterion is set as 1.0E–3. If the calculation converges, the calculation process then returns to part I to calculate the pressures and mass flow rates of nodes again. Before returning to part I, the adjusted flow coefficient, which is established to simplify the calculation of mass flows, is updated to reflect the energy transfer on the mass flows. The adjusted flow coefficient is represented as C ¢ = ρi CA( i , j )
3 3.1
Verification of CAU_ESCAP Without energy transfer
A new network-based smoke control program, CAU_ESCAP, is developed to estimate the smoke movement with consideration for the energy transfer in ultra-high-rise buildings. The result of smoke control systems on the area of refuge (AOR) (Klote 1993) is used for validating CAU_ESCAP when the energy transfer is ignored. Figure 3 shows the floor plan of an 11-story building, and it is assumed that all doors between the office and the corridor are opened. Table 1 lists the leakage areas and the conditions of simulation (Klote 1993). In CAU_ESCAP, the smoke movement in the building is calculated with the assumption that the fire and the smoke control system are in steady-state conditions, so the fire is regarded as fully developed. Generally, in a fully developed fire, flashover would occur in the fire room, and the windows in the fire room would then be broken by the fire, such that the inflow and outflow of the air occur freely through the broken windows. Therefore, it is assumed that the windows in the fire floor are all opened for considering this phenomenon in the steady state. Also, as listed in Table 1, the temperature of the fire room is set to 873.15 K, and the mass flow rate of the combustion gas generated by the fire is set to 0.6316 kg/s. The pressurization of an elevator shaft is selected for protecting the infiltration of smoke in the area of refuge, and the mass flow rate of the smoke control system is set to 5.4 kg/s (Klote 1993). Table 2 lists the pressure difference between the office and the area of refuge from the 2nd to the 11th floors. As listed in Table 2, there is no difference between CAU_ESCAP and ASCOS, when neglecting the effect of energy transfer from the outdoor temperature to the building. Figure 4 shows the distribution of pressure in the building. Because all of the pressures in the building are higher than the outside pressure due to pressurization of the elevator shaft, the infiltration of combustion gas from the fire room to other spaces in the building does not occur. As listed in Table 3, the temperature of the fire room is estimated as
(11)
where, C is the flow coefficient, generally taken to be in the range of 0.6 to 0.65 (Klote 1991). A( i , j ) is the leakage area between nodes i and j, and ρi is the density of air calculated according to temperature.
Fig. 3 The floor plan of the building for validation
177
Bae et al. / Building Simulation / Vol. 6, No. 2 Table 1 Leakage areas between each compartment and boundary conditions for analysis (Klote 1993) Leakage area for each compartment Leakage area (m2)
Location
Condition Property
Value
1st floor (opened door)
0.975
Building temperature
294.15 K
Others floor (closed windows)
0.0204
Outside temperature
294.15 K
Others floor (opened windows)
1.86
Fire temperature
873.15 K
Stair to building (closed door)
0.0251
Fire floor
11th floor
Building floor
0.0204
Wind velocity
0 m/s
Building to AOR (closed door)
0.0390
Mass loss rate
0.6316 kg/s
AOR to hoistway (closed door)
0.149
Smoke control system
Pressurizing elevator : 5.4 kg/s
Table 2 Pressure difference (Pa) for case 1 between AOR and an office from the 2nd to the 11th floor 2 ASCOS
3
4
5
6
7
8
9
10
11
33.1 31.1 30.8 30.7 30.7 30.7 30.8 31.1 33.2 81.2
CAU_ESCAP 33.0 31.0 30.7 30.6 30.6 30.6 30.6 31.0 33.0 81.3
Table 3 Temperature (K) distribution from the 2nd to the 11th floor ASCOS
CAU_ESCAP
Floor Office AOR Stairwell Elevator Office AOR Stairwell Elevator 2
294.15 294.15 294.15
294.15 294.15 294.15 294.15
294.15
3
294.15 294.15 294.15
294.15 294.15 294.15 294.15
294.15
4
294.15 294.15 294.15
294.15 294.15 294.15 294.15
294.15
5
294.15 294.15 294.15
294.15 294.15 294.15 294.15
294.15
6
294.15 294.15 294.15
294.15 294.15 294.15 294.15
294.15
7
294.15 294.15 294.15
294.15 294.15 294.15 294.15
294.15
8
294.15 294.15 294.15
294.15 294.15 294.15 294.15
294.15
9
294.15 294.15 294.15
294.15 294.15 294.15 294.15
294.15
10
294.15 294.15 294.15
294.15 294.15 294.15 294.15
294.15
11
873.15 294.15 294.15
294.15 494.66 294.15 294.15
294.15
Fig. 4 Pressure distribution of a building calculated by CAU_ESCAP
3.2
With energy transfer
It is necessary to verify the new network-based smoke control program, CAU_ESCAP, because its suitability has not been verified when the calculation of the energy transfer is applied. The distributions of pressure and temperature in a high-rise building of 40 stories (Black 2010) are calculated for validating the calculation of the energy transfer. As mentioned in the introduction, the fire is expressed by the pressure and temperature in the COSMO program, and it is necessary to alter the pressure into the mass flow rate of the source for simulating with CAU_ESCAP. As listed in Table 4, Black (2010) assumed that the pressure rise on the fire floor is 10 Pa above the ambient pressure, and that the temperature of the combustion gas is 973.15 K. The pressure rise of 10 Pa above the ambient pressure can then be altered into a mass flow rate of the source of 0.1205 kg/s. Table 4 lists the specific fire conditions and building scenario (Black 2010). Table 4 Input conditions for analysis (Black 2010) Input variable
494.66 K with CAU_ESCAP, which is lower than the fire temperature, 873.15 K, while ASCOS predicts the same temperature as the fire. This is a result of mixing with the pressurization air (294.15 K) and by the release of energy from the fire room to the outside. Therefore, the energy transfer is neglected in the entire calculation, while all of the energy generated from the fire is released to the outside. Thus, from this result, it is verified that the loop for calculating the enthalpy conservation does not affect the whole calculation of the CAU_ESCAP, when the energy transfer is ignored.
Height of each floor
COSMO
CAU_ESCAP
4m
4m
Area of vent at top (elevator)
10.224 m2
10.224 m2
Area of vent at top (stairwell)
0.1 m2
0.1 m2
0.0097 m
2
0.0097 m2
Leakage area with floor (stairwell)
0.0026 m
2
0.0026 m2
Surface temperature (stairwell)
313.15 K
313.15 K
Surface temperature (elevator)
313.15 K
313.15 K
Temperature of room
295.15 K
295.15 K
Temperature of ambient
256.15 K
256.15 K
Temperature of fire
973.15 K
973.15 K
10 Pa
0.1205 kg/s
Leakage area with floor (elevator)
Pressure rise/source of mass flow rate
178
As shown in Fig. 5(a), the pressure distribution of the room from CAU_ESCAP is the same as the result from COSMO. However, there is an inappreciable difference in the pressure distribution for the stairwell and the elevator, which is caused by the difference in the temperature distributions between CAU_ESCAP and COSMO. As shown in Fig. 5(b), the tendency of the temperature distributions is similar between CAU_ESCAP and COSMO. With CAU_ESCAP, the temperature in the stairwell is estimated higher than in the elevator shaft along the shaft to the top, while COSMO estimates the temperature in the elevator shaft more highly than in the stairwell. This difference in the temperature distributions appeared due to the difference between the treatment of the boundary for the energy transfer, and this causes the differences in the pressure distributions in the shaft. However, as listed in Table 4, the area of the vent at the top of the elevator shaft, 10.224 m2, is 100 times larger than that of the stairwell, which is 0.1 m2. Therefore the amount of accumulated energy in the elevator shaft released to the outdoors is much higher than that in the stairwell. With this phenomenon, it is verified that the energy calculation of CAU_ESCAP is quite suitable for analyzing the smoke movement phenomenologically. 4 Results and discussions Using CAU_ESCAP, an ultra-high-rise building fire is simulated for analyzing the smoke movement and
Bae et al. / Building Simulation / Vol. 6, No. 2
determining the mass flow rate of a smoke control system. Figure 6 shows the schematic of an ultra-high-rise apartment with 55 stories applied for simulation. It is assumed that all doors between the vestibules and room are opened, and the building space on each floor can be modeled as one node. The fire is assumed to be fully developed, and that the window in the fire room would be broken, as previously mentioned (Quintiere 1998). In general, the temperature of the fire room is between 1073.15 K and 1273.15 K for a fully developed fire (Quintiere 1998). Therefore, the fire temperature is assumed to be 1173.15 K, and PMMA is selected as the fuel of the fire source because it is widely used for electronic products. Table 5 lists the leakage areas between each compartment, and the conditions of the fire and boundaries. The conditions of the leakage areas are from the values of Seo et al. (2010), 0.01 m2. Generally, the qualification for pressure difference between a fire room and a protected area is from 40 to 60 Pa (NFPA 2000; Lee et al. 2007; IFC 2009). Then, at the smoke protecting area, the inflow of smoke has to be protected, and the door opening and closing have to be smooth for all occupants in the building. In this study, the stairwell is selected as the smoke protecting area, and then the total mass flow rate of smoke control are estimated for each fire case and the conditions of the door on the fire floor, as listed in Table 5. It is assumed that the fire occurred on the 2nd floor, because the fire on the lowest floor is the worst case for evacuating occupants in buildings. The fire
Fig. 5 Distributions of the pressure and temperature for both CAU_ESCAP and COSMO: (a) pressure distribution; (b) temperature distribution
179
Bae et al. / Building Simulation / Vol. 6, No. 2
Fig. 6 Building of the example analysis of fire situation/smoke control system Table 5 Leakage areas between each node, and fire and boundary conditions for analysis Leakage area for each compartment Location
Leakage area
Room: outside
0.01 m2
Room: outside (broken window)
1.8 m2
Room: stairwell
0.01 m2
Room: stairwell (opened condition)
2.1 m2
Fire condition Condition Fire position
Fire size
Other boundary condition
Value
Property
Condition
2nd floor
Building & outside temperature
274.15 K
3 MW Wind velocity
0 m/s
5 MW Floor height
3.0 m
10 MW
Smoke control system
Pressurizing stairwell
size is set to 3 MW, 5 MW and 10 MW, considering the maximum heat release rate of representative combustibles in living spaces (Quintiere 1998). Figure 7 shows the distributions of the pressure difference between each node and the temperature of the stairwell. Even though the heat release rate increases, the distributions of the pressure difference and the stairwell temperature are the same for all cases, when the door between the building and stairwell is closed. As listed in Table 6, there is no
variation of mass flow rate from the fire room to the stairwell, regardless of heat release rate (Appendix, Table A1). However, the mass flow rate from the fire room to the exterior is increased. The source flows from the fire mainly flow out to the exterior due to the difference in leakage area between a broken window (1.8 m2) and a closed door (0.01 m2). Therefore, the effect of the heat release rate is negligible for the distributions of pressure difference and the stairwell temperature. Furthermore, the distribution of stairwell temperature is similar to the shape of an exponential function, since a small quantity of air flows into the stairwell through the small leakage area between the upper floor of the fire room and the lower floor of the neutral plane, and is mixed with hot combustion gas from the fire. Figure 8 shows the distributions of the pressure difference and the stairwell temperature, in the case of an opened door. The distributions of the pressure difference and the stairwell temperature are increased as the heat release rate increases. The distribution of the temperature in the stairwell is estimated as lower than the fire temperature, at 1173.15 K, and the estimated temperature does not vary with the height in the case of a closed door. Because the combustion gas at high temperature (1173.15 K) is mixed with the outside air at low temperature (294.15 K) by substantial inflows, the temperature in the fire room is decreased. Also, the combustion gas mixed with outside air flows into the stairwell from the fire room then exits the stairwell through the building compartment above the neutral plane (Table 7). Therefore, the temperatures in the stairwell are estimated as the same value between the 2nd floor and the 54th floor. Moreover, as listed in Table 7, the height of the neutral plane for an opened door of the stairwell is predicted as much lower than for the closed case. This result means that the strength of the stack effect for an opened door of the stairwell is stronger than the closed case, and the stairwell would be filled with smoke from the fire and be unsuitable as an evacuation route if there is no smoke control system. Table 8 lists the pressure difference and the fire protection velocity for several mass flow rates of pressurization. In the case of a closed door of the stairwell, the pressure difference between the fire room and the stairwell is adjusted to 50 Pa, the standard intermediate value for the smoke protection. In the case of an opened door, the fire protection velocity for the safety region is adjusted to 1 m/s, as suggested by IFC (2009). Because the object building described in this paper has high air tightness, the qualifications of the pressure difference and the smoke protection velocity would be satisfied with a small amount of mass flow rate for pressurization. However, in the case of an opened door of the stairwell, the mass flow rate of the pressurization increases to 5.5%, 6.9%, and 9.7% for each heat release rate. Particularly, a variation
180
Bae et al. / Building Simulation / Vol. 6, No. 2
Fig. 7 The distributions of the pressure difference and the temperature between a building compartment and the stairwell when the door in the fire room is closed
Fig. 8 The distributions of the pressure difference and the temperature between a building compartment and the stairwell when the door in the fire room is opened Table 6 The mass flow rate (kg/s) around the fire room Closed stairwell door
Opened stairwell door
3 MW
5 MW
10 MW
3 MW
5 MW
10 MW
–0.0615
–0.0616
–0.0615
–3.0937
–3.5479
–4.1848
To outside –0.0590
–0.1393
–0.3401
2.9732
3.3471
3.7831
To stairwell
Table 7 The positions of the neutral plane for each case Closed stairwell door 3 MW Position of neutral 79.43 m plane
Opened stairwell door
5 MW
10 MW
3 MW
5 MW
10 MW
79.42 m
79.42 m
10.53 m
10.35 m
10.01 m
Table 8 The pressure difference and the fire protection velocity for different mass flow rates Closed stairwell door
Opened stairwell door
Pressure difference between fire Smoke protection velocity Mass flow room and stairwell (Pa) from fire room (m/s) rate of SCS (kg/s) 3 MW 5 MW 10 MW 3 MW 5 MW 10 MW 0.0475
50
49.8
49.8
0.9476
0.9333
0.9048
0.0476
50.1
50.0
50.0
0.9476
0.9333
0.9048
0.0501
55.5
55.3
55.3
1.0
0.9857
0.9571
0.0508
57.1
56.9
56.9
1.0143
1.0
0.9714
0.0521
60.2
60.0
60.0
1.0381
1.0286
1.0
Bae et al. / Building Simulation / Vol. 6, No. 2
of the pressure difference occurs in the acoustic velocity (Lee et al. 2011), so the smoke from the fire would be spread to other spaces if there is no instantaneous increase in the mass flow rate for pressurization. 5 Conclusions A new network-based smoke control program, CAU_ESCAP, has been developed in this study to simulate a fire with energy transfer between each node. CAU_ESCAP has been validated with previous programs, ASCOS and COSMO, by analyzing the smoke movement in the building. Moreover, fire in an ultra-high-rise building of 55 stories was applied in CAU_ESCAP for analyzing the smoke movement and determining the mass flow rate of the smoke control system. The conclusions of this study are as follows: (1) The pressure in the building is enormously influenced by the temperature of each node, so the energy transfer has to be considered to analyze the smoke control and pressurization. (2) CAU_ESCAP has been validated with previous programs, ASCOS and COSMO. It is appropriate for the analysis of fire situations due to the inclusion of energy transfer. (3) The height of the neutral plane for an opened door of the stairwell was estimated as much lower than for the closed case. This means that the strength of the stack effect for an opened door of the stairwell is stronger than the closed case, and the stairwell would be filled with smoke from the fire and be unsuitable as an evacuation route if there is no smoke control system. (4) The pressure difference does not vary with the heat release rate of the fire, but it varies with the condition of the door between the fire room and the protecting area. Variation of the pressure difference occurs in the acoustic velocity with changes to the door condition, so the smoke from the fire would be spread to other spaces if there is no instantaneous increase in the mass flow rate for pressurization. Acknowledgements This research was supported by a grant (Code# ’09 R&D A01, Supertall Building R&BD Center) from Cutting-edge Urban Development Program funded by Ministry of Land, Transport and Maritime Affairs, Republic of Korea. References Black WZ (2009). Smoke movement in elevator shafts during a high-rise structural fire. Fire Safety Journal, 44: 168 182.
181 Black WZ (2010). COSMO·Software for designing smoke control systems in high-rise buildings. Fire safety Journal, 45: 337 348. Chapra SC, Canale RP (2005). Numerical Methods for Engineers, 5th edn. New York: McGraw-Hill. Choi Y (2010). Haeundae Woosin Gonden Sweet Fire! Fire Prevention News. (in Korean) IFC (2009). International Fire Code, Section 909: Smoke control systems. Washington, DC: International Code Council. Klote JH (1991). Design manual for smoke control systems, NISTIR 4551. Gaithersburg, USA: National Institute of Standards and Technology. Klote JH (1993). Design of smoke control systems for areas of refuge. ASHRAE Transactions, 99(2): 793 807. Klote JH (1995). An overview of smoke control research. ASHRAE Transactions: Symposia, 101(1): 979 990. Klote JH, Bodart X (1985). Validation of network models for smoke control analysis. ASHRAE Transactions, 91(2B): 1134 1145. Klote JH, Fothergill JW (1983). Design of Smoke Control Systems for Buildings. Atlanta: American Society of Heating, Refrigerating and Air-Conditioning Engineers. Lee DM, Park SM, Cho YS, Park HS (2007). NFSC 501A: A manual for smoke control systems in the stairwells and vestibules of specific fire escape stairs. Saerombooks. (in Korean) Lee MJ, Kim NI, Ryou HS (2011). Air tightness measurement with transient methods using sudden expansion from a compressed chamber. Building and Environment, 46: 1937 1945. Ministry of Land (2009). Statistics of Constructions. Daejeon, R.O. Korea: Statistics Korea. (in Korean) NFPA (2000). NFPA 92A: Recommended practice for smoke control systems. Quincy, USA: National Fire Protection Association. Park JC (2009). Standard of smoke control systems in ultra-high-rise buildings. Magazine of SAREK, 38(11): 10 25. (in Korean) Quintiere JG (1998). Principles of Fire Behavior. New York: Delmar Publishers. Schwarz A, Janicka J (2009). Combustion Noise. New York: Springer. Seo BY, Choi JH, Hong WH (2010). Prediction of smoke diffusion and minimizing methods of stack effect considering the status of opening in a high-rise building. Architectural Institute of Korea, 26(9): 259 266. (in Korean) Sonntag RE, Borgnakke C, Van Wylen GJ (2005). Fundamentals of Thermodynamics, 6th edn. New York: John Wiley & Sons. Walton GN (1989). AIRNET—A Computer Program for Building Airflow Network Modeling. Gaithersburg, USA: National Institute of Standards and Technology. Walton GN (1994). CONTAM93 User Manual. Gaithersburg, USA: National Institute of Standards and Technology. Wang L, Dols WS, Chen Q (2011). Using CFD capabilities of CONTAM 3.0 for simulating airflow and contaminant transport in and around buildings. HVAC & Research, 16: 749 763. White FM (2008). Fluid Mechanics, 6th edn. New York: McGraw-Hill. Yu MY, Cha KS, Park MS (2007). Propagation route of smoke at the examples of ultra high-rise building fire in North America. Magazine of SAREK, 36(2): 33 40. (in Korean)
182
Bae et al. / Building Simulation / Vol. 6, No. 2
Appendix Table A1 The mass flow rate (kg/s) with the door closed without smoke control Floor 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55
3 MW –0.05900 0.08280 0.07910 0.07575 0.07264 0.06972 0.06695 0.06427 0.06167 0.05913 0.05662 0.05413 0.05165 0.04915 0.04663 0.04406 0.04142 0.03868 0.03583 0.03280 0.02953 0.02594 0.02183 0.01682 0.00956 –0.00935 –0.01598 –0.02056 –0.02429 –0.02751 –0.03040 –0.03303 –0.03548 –0.03776 –0.03989 –0.04193 –0.04387 –0.04573 –0.04751 –0.04923 –0.05089 –0.05249 –0.05405 –0.05556 –0.05703 –0.05847 –0.05987 –0.06123 –0.06256 –0.06387 –0.06515 –0.06641 –0.06764
–0.06151 –0.08280 –0.07910 –0.07574 –0.07264 –0.06972 –0.06694 –0.06426 –0.06167 –0.05912 –0.05662 –0.05413 –0.05165 –0.04915 –0.04663 –0.04405 –0.04141 -0.03868 –0.03582 –0.03279 –0.02953 –0.02593 –0.02183 –0.01682 –0.00954 0.00937 0.01599 0.02057 0.02430 0.02752 0.03040 0.03304 0.03548 0.03776 0.03990 0.04193 0.04387 0.04573 0.04751 0.04923 0.05089 0.05250 0.05405 0.05557 0.05704 0.05847 0.05987 0.06123 0.06256 0.06387 0.06515 0.06642 0.06765
5 MW 0.12715 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 –0.09737
–0.13929 0.08285 0.07915 0.07580 0.07269 0.06977 0.06699 0.06431 0.06171 0.05917 0.05666 0.05417 0.05168 0.04919 0.04666 0.04409 0.04144 0.03871 0.03584 0.03282 0.02954 0.02595 0.02183 0.01682 0.00953 –0.00938 –0.01600 –0.02059 –0.02431 –0.02753 –0.03043 –0.03306 –0.03550 –0.03778 –0.03993 –0.04196 –0.04390 –0.04576 –0.04754 –0.04926 –0.05092 –0.05253 –0.05409 –0.05560 –0.05707 –0.05851 –0.05991 –0.06127 –0.06260 –0.06391 –0.06519 –0.06644 –0.06767
–0.06155 –0.08285 –0.07915 –0.07580 –0.07269 –0.06977 –0.06699 –0.06431 –0.06171 –0.05916 –0.05666 –0.05417 –0.05168 –0.04918 –0.04666 –0.04408 –0.04144 –0.03870 –0.03584 –0.03281 –0.02954 –0.02594 –0.02183 –0.01682 –0.00954 0.00939 0.01601 0.02059 0.02431 0.02754 0.03043 0.03306 0.03550 0.03778 0.03993 0.04196 0.04390 0.04576 0.04755 0.04927 0.05092 0.05254 0.05409 0.05561 0.05708 0.05851 0.05991 0.06127 0.06260 0.06391 0.06519 0.06644 0.06767
10 MW 0.12723 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 –0.09741
–0.34015 0.08278 0.07908 0.07573 0.07262 0.06970 0.06692 0.06425 0.06165 0.05910 0.05660 0.05411 0.05163 0.04913 0.04661 0.04404 0.04140 0.03866 0.03580 0.03278 0.02951 0.02591 0.02180 0.01680 0.00952 –0.00939 –0.01599 –0.02058 –0.02430 –0.02752 –0.03041 –0.03304 –0.03548 –0.03776 –0.03990 –0.04193 –0.04387 –0.04573 –0.04751 –0.04923 –0.05089 –0.05250 –0.05405 –0.05557 –0.05703 –0.05847 –0.05987 –0.06122 –0.06255 –0.06386 –0.06514 –0.06639 –0.06762
–0.06153 –0.08278 –0.07908 –0.07573 –0.07263 –0.06971 –0.06693 –0.06425 –0.06165 –0.05911 –0.05660 –0.05411 –0.05163 –0.04914 –0.04662 –0.04404 –0.04140 –0.03867 –0.03581 –0.03279 –0.02952 –0.02593 –0.02182 –0.01681 –0.00953 0.00938 0.01599 0.02056 0.02429 0.02751 0.03040 0.03303 0.03547 0.03775 0.03989 0.04193 0.04387 0.04572 0.04750 0.04922 0.05089 0.05249 0.05405 0.05556 0.05703 0.05846 0.05986 0.06122 0.06255 0.06386 0.06514 0.06639 0.06762
0.12713 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 –0.1309
