Fire safety in heritage structures: A case study on heritage temple with performance-based design fire approach

doi:10.21203/rs.3.rs-3621257/v1

Heritage structures are continuously subject to fire accidents due to the presence of combustible materials that result in the loss of heritage as well as the fatality of occupants. The probability of fatalities increases when the limit of the number of occupants is violated. Fire accidents in Indian heritage temples witnessed a huge number of deaths due to incapacitation and stampedes. This study aims to evaluate fire hazards (thermal, visibility, and toxic hazards) using a performance-based design fire approach for an Indian heritage temple. A heritage model temple was created in Fire Dynamics Simulator (FDS) with dimensions of 19.25 m × 19.25 m × 9.75 m. A design fire of 5 MW was specified for three different fire growth rates i.e., medium (0.012 kW/s²), fast (0.047 kW/s²), and ultrafast (0.19 kW/s²). Thermal, visibility, and toxic hazards expressed in terms of temperature, visibility, and concentration of CO₂ and CO reached the tenable limit in 170, 160, 383, and 176 s respectively. Thus Available Safe Egress Time (ASET) for occupants was only 160 s. The effect of soot yield and ventilation conditions on fire hazards was also analyzed. Required Safe Egress Time (RSET) estimated with an egress model for an occupant load factor of 1.4 m²/ person as per the National Building Code of India 2016 was found to be 109 s. The safety margin between ASET and RSET was increased significantly with the help of two smoke extractors. The results obtained from this study will provide valuable insights into fire safety measures in similar heritage structures.

Heritage

Temple

Fire hazards

ASET

RSET

Stampede

Fire is one of the major causes of damage to heritage buildings and monuments around the world (Garcia-Castillo, Paya-Zaforteza and Hospitaler, 2021). Fire could result from electric short circuits, smoking, hot works, eternal flame, and lightning (Fire risk in heritage properties, 2021). Heritage buildings must be protected from fire due to their historical and socio-cultural significance. There are several reports of heritage buildings being destroyed, including temples, due to fire. In particular, heritage buildings with high fire loads and a lack of adequate safety measures are more prone to fire damage. The fire accidents in Indian heritage temples such as Puttingal Temple, Meenakshi Amman Temple, and Mulamkadakam Temple manifest the lack of fire safety measures (Revenue and Disaster Management, 2016; Govindrajan, 2018; The Hindu, 2021). In recent years, some of the world’s most renowned heritage destruction by fire were the National Museum of Brazil, Shuri Temple, and Notre Dame Cathedral (Mega, 2019; Ball, 2020; Jiji, 2020). Fire accidents at heritage buildings lead to the loss of invaluable objects, knowledge, and sometimes the whole building. Additionally, if crowd control is poorly managed in areas with a large number of visitors, the likelihood of fatality increases due to incapacitation and stampedes (Bernardini, 2017; Vanumu, Laxmikant and Rao, 2020).

The use of prescriptive codes like the National Building Code of India 2016, Part IV (Bureau of Indian Standards, 2016) is necessary to be implemented. However, there are also challenges in evaluating fire safety using the prescriptive approach for heritage buildings. As they are already built-in, the provision of structural changes could be minimal to maintain the aesthetics of heritage (Marrion, 2016). Also, the National Building Code of India 2016, Part IV does not provide any information regarding the tenability. Therefore, some scholars have underlined the importance of performance-based fire safety evaluations of heritage buildings (Bukowski and Nuzzolese, 2009; Torero, 2019). It is the key responsibility of the government and the subsequent ruling authorities to analyze the fire safety situation of the heritage buildings and thus implement the available standard codes and practices, however, the analysis of fire performance-based analysis is pivotal. Performance-based fire design is a novel strategy that prioritizes meeting specific fire safety goals over complying with restrictive building standards. Using a performance-based approach, fire prevention measures can be tailored based on the detailed examination of a building's characteristics, fire dynamics, and occupant behavior.

The use of Computational Fluid Dynamics (CFD) based software has increased in recent years as a performance-based tool to analyze the fire safety situation in the buildings. Considerable research with the variety of fire scenarios for small spaces, high-rise buildings, oil and gas facilities, etc. (Liu et al., 2015; Long, Zhang and Lou, 2017; Betting et al., 2019; Lim et al., 2019; Khan, Usmani and Torero, 2021; Kumar et al., 2022) have been simulated using Fire Dynamics Simulator (FDS), however, very few studies are available for heritage buildings (Donghyun et al., 2020; Su et al., 2020; Huai et al., 2021). FDS is a CFD software for low-speed thermally driven flow from fire, developed by the National Institute of Standards and Technology (NIST), USA, and VTT Technical Research Centre of Finland (McGrattan et al., 2013). FDS is a very well-verified and validated software that gives reliable results (Wen et al., 2007; K McGrattan et al., 2013a, 2013b; Van Hees, 2013). Our group has also validated FDS simulations with experimental results (Kumar et al., 2022). The studies using FDS have evaluated the following parameters: temperature, visibility, and concentration of combustion gases. These parameters are regarded as thermal, visibility, and toxic hazards used for evaluating fire safety and tenability conditions.

This paper tries to portray the fire and smoke spread in a typical Indian heritage temple using numerical simulation with the help of FDS software. The study aimed to predict the fire hazards, evaluate the tenability condition, and the best time for occupants to escape the temple. The sensitivity of various model parameters that affect the intensity of fire and smoke like fire growth coefficient, soot yield, and ventilation condition were evaluated. Thermal, toxic, and visibility hazards were determined to find out ASET. The egress model was developed coupled with the fire dynamics model to obtain RSET. Also, the effect of smoke extractors in improving the tenability was analyzed.

2.1. Details of the temple geometry and boundary conditions

The structure considered for the simulation of fire and smoke is a typical centuries-old Kalyana Mandapa of Vijaya Vittala temple, Hampi with some modifications (Natampally, 2018). The temple covers a floor area of 230 m² with dimensions 19.25 m $\times$ 19.25 m $\times$ 9.75 m, consisting of a hall with a ceiling height of 4.45 m from the floor which is at 1.5 m from the ground with three doorway openings of size 1.5 m $\times$ 3.5 m. The temple geometry was created in the Fire Dynamics Simulator (FDS). The plan, elevation, side view, and isometric view are shown in Fig. 1(a-d). FDS utilizes Navier-Stokes equations for thermally driven flow and uses Large Eddy Simulation (LES) as the turbulence model (McGrattan et al., 2013). The LES technique assumes that the numerical mesh is fine enough to allow for the formation of eddies. The fundamental equations for FDS to run the simulation are continuity (1), momentum transport (2), energy conservation (3), and species conservation (4) (McGrattan et al., 2013). The simulation was performed on the hardware platform of Intel Xeon CPU E3-1270 V5 with a speed of 3.60 GHz and 32 GB RAM. The details of the simulation model are given in Table 1.

$$\frac{\partial {\rho }}{\partial \text{t}}+\nabla .{\rho }\overrightarrow{\text{v}}=0 \left(1\right)$$

$${\rho }\left(\frac{\partial \overrightarrow{\text{v}}}{\partial \text{t}}+\frac{1}{2}\nabla {\left|\overrightarrow{\text{v}}\right|}^{2}\right)+\nabla \text{P}-{\rho }\text{g}={\overrightarrow{\text{f}}}_{\text{b}}+\nabla .{{\tau }}_{\text{i}\text{j}} \left(2\right)$$

$$\frac{\partial }{\partial \text{t}}\left({\rho }\text{h}\right)+\nabla .\left({\rho }\text{h}\overrightarrow{\text{v}}\right)=\frac{\partial \text{P}}{\partial \text{t}}+\overrightarrow{\text{v}}.\nabla \text{P}-\nabla {\overrightarrow{\text{q}}}_{\text{r}}+\nabla .\left(\text{k}\nabla \text{T}\right)+\sum _{\text{i}}\nabla \left({\text{h}}_{\text{i}}{\rho }{\text{D}}_{\text{i}}\nabla {\text{Y}}_{\text{i}}\right) \left(3\right)$$

$$\frac{\partial }{\partial \text{t}}\left({\rho }{\text{Y}}_{\text{i}}\right)+\nabla .\left({\rho }{\text{Y}}_{\text{i}}\overrightarrow{\text{v}}\right)=\nabla \left({\rho }{\text{D}}_{\text{i}}\nabla {\text{Y}}_{\text{i}}\right)+{\dot{\text{m}}}_{\text{i}}^{{\prime }{\prime }{\prime }} \left(4\right)$$

2.2. Description of fire scenario

A fire source of 2 $\times$ 2 m² was created at the center of the temple with a maximum heat release rate (HRR), also called heat flux, equal to 5 MW (Fig. 2). In the real scenario, the fire was assumed to start from the eternal flame present near the idol(s) which were assumed to be placed at the center of the temple. The combustible materials that can be possibly present in a temple include wood, upholstery, plastics, ghee/liquid fuel, paper, and flowers. Three fire scenarios were considered based on varying fire growth rates i.e., medium (0.012 kW/s²), fast (0.047 kW/s²), and ultrafast (0.19 kW/s²) as per the Society of Fire Protection Engineering (SFPE) guidelines for different categories of combustibles (Morgan J. Hurley, 2016). The scenario with ultrafast fire was considered to be the worst-case scenario. The measurement of the hazards like temperature, the concentration of gases (CO₂ and CO), and visibility were done with the help of devices like thermocouples (T), gas detectors (CO₂ and CO), visibility detectors (V), and smoke detectors (SD) (Fig. 2). Smoke detectors were placed 5 m apart (< 9.1 m) and 200 mm below the ceiling (< 300mm) as per NFPA 72 (National Fire Protection Association, 2019). As per the National Building Code of India (2016), for the occupancy type of less concentrated assembly without seating, an occupant density load factor of 1.4 m²/person was considered (Bureau of Indian Standards, 2016). The occupants were assumed to escape the temple from the doorways with an average velocity of 1.2 m/s as per the SFPE guide for the considered occupancy load factor (O’Connor, 2019).

2.3. Methodology

The simulation study was conducted in several stages. Initially, grid sensitivity analysis was performed to optimize the mesh size. In the second stage, fire and smoke spread with time were simulated, and fire hazards (thermal, visibility, and toxic hazards) were evaluated for different fire growth coefficients. The time for each fire hazard to reach the tenability limit was evaluated and compared with the tenability criteria to estimate the ASET. In the third stage, the effect of soot yield and sensitivity of ventilation conditions on fire hazards were analyzed. The variations in fire growth coefficient, soot yield, and ventilation condition are shown in Table 2. Then, RSET was evaluated with the egress model and compared with ASET. Finally, the effect of smoke exhaust on tenability is shown.

Table 1

Details of the Simulation Model
Model quantities	Description
Model size	19.25 m $\times$ 19.25 m $\times$ 9.75 m
Size of the fire source	4 m² (2 m $\times$ 2 m)
Vent size	1.5 m $\times$ 3.5 m approx.
Number of vents	1–3 doorway
Size of mesh in rest of the model	0.25 m
Size of mesh around fire source	0.0625–0.125 m
Heat of combustion	20 MJ/kg
Radiative fraction	0.35
HRRPUA	1.25 MW/m²
Max HRR	5 MW
Fire growth coefficient (α)	0.012 (medium) – 0.19 (ultrafast) kW/s²
Soot yield	0.02–0.10 g/g_fuel
CO yield	0.05 g/g_fuel
Simulation time	600 s
Simulation type	LES
Ambient temperature	20°C
Ambient pressure	1 atm

Table 2

Variation of Model Parameters for Sensitivity Analysis
Simulation no.	Fire growth coefficient α (kW/s²)	Number of doorways	Soot yield (g/g_fuel)
1.	0.012	2	0.06
2.	0.047	2	0.06
3.	0.19	2	0.06
4.	0.19	1	0.06
5.	0.19	2	0.06
6.	0.19	3	0.06
7.	0.19	2	0.02
8.	0.19	2	0.06
9.	0.19	2	0.10

3.1. Grid sensitivity analysis

Initially, simulation was performed with 25 cm mesh which did not produce fire and smoke. Therefore, the mesh was reduced to 12.5 cm around the fire source allowing fire and smoke to be produced. As per earlier studies (Wen et al., 2007), the predicted results could underestimate or overestimate the actual fire if the grid size is coarse. The deviations can be reduced with the mesh refinement. However, the refinement in mesh size greatly increases the computation time. For instance, in the present case, the time to simulate the coarse (12.5 cm), medium (8.33 cm), and fine (6.25 cm) mesh was approximately 44, 66, and 112 h respectively. Considering the small deviation in temperature and large difference in computation time with respect to the mesh size (Table 3 & Fig. 3d), medium mesh was chosen as the optimal mesh for further investigation.

Table 3

Grid optimization for 600 s simulation
Mesh	Coarse	Medium	Fine
Size (cm)	12.5	8.33	6.25
No. of cells	78,848	2,66,112	6,30,784
Computational time (h)	44	66	112

3.2. Fire and smoke development

The development of fire and smoke with medium mesh size, ultrafast fire growth (worst-case), two doorways, and 0.06 g/g_fuel soot yield to understand the changes in fire hazards with time. The fire hazards were computed at 1.6 m (human height) at two locations: the center of the temple (donated as ‘c’ in graphs) and near the doorway (donated as ‘d’ in graphs). The visualization of fire and smoke development is shown in Fig. 4. The figure depicts both the interior and exterior of the temple at various times. It can be seen that fire and smoke spread rapidly which is due to high HRR value and ultrafast fire growth. This was done to illustrate the worst-case fire scenario that could occur inside a temple of this size. The movement of smoke was as follows: smoke was observed to rise initially to the ceiling and then started to decrease when enough smoke accumulated at the top. Finally, smoke started to disappear from the lower height when exuded from the doorways.

The size of the fire depends upon the characteristics and density of the fire load. The size of the fire source actually affects the amount of smoke produced and the rate at which heat is released when burning advances over time. A "t-squared fire" depicts the growth rate of a design fire and demonstrates how time squared affects the rate of heat release. The t-squared parabolic growth equation is given by $Q =\alpha .{t}^{2}$ where Q is the HRR (kW), $\alpha$ is the fire growth coefficient (kW/s²), and t is time (s). In general, the fire growth rate is divided into four categories: slow ($\alpha =0.003$), medium ($\alpha =0.012$), fast ($\alpha =0.047$), and ultrafast ($\alpha =0.19$). The latter three were selected to find the sensitivity of fire growth to fire hazards. Different coefficients were selected as per the Society of Fire Protection Engineering (SFPE) guidelines for different categories of combustibles (Morgan J. Hurley, 2016). The effect of the fire growth coefficient on HRR and fire hazards is shown in Fig. 5 with a tenability line (horizontal dashed line). Also, the time to reach untenability for each fire hazard at human height is shown with the help of visualization slices in Fig. 6. The slices were made to pass through the center of the fire and two doorway openings. We verified that the most vulnerable situation was present where the slice was made.

3.2.1. Temperature rise at different locations

Exposure to high temperature affects the occupants in three ways: hyperthermia, skin burns, and respiratory tract burns. The temperature at the ceiling height (4.3 m) in all the locations was higher than the human height (1.6 m) which must be due to the rise of smoke and hot gases and their accumulation near the ceiling. The temperature was highest at the center of the temple among the two computed locations which is obvious for being the fire location. It was found that the temperature near doorways was higher than the temperature near the walls. It must be due to convective heat transfer and temperature gradient created from inside (higher temperature area) to outside (lower temperature area). Thermal burns of the respiratory tract could occur when air above 60°C is inhaled (Gager III and Dominguez, 2016). If we consider only the human height for ultrafast fire (Fig. 5b), the unbearable temperature of 60°C reached the fastest at the center in only 50 seconds whereas it took about 170 seconds to reach the same level at the doorway. Thus the occupants have only 170 seconds (less than 3 min) to escape from the temple. This can be verified with the slice in Fig. 6a. The higher temperature at the center might not affect the occupants instantaneously because all the hot smoke rises vertically first rather than spreading horizontally.

3.2.2. Visibility

In the fire process, visibility is the most critical risk element as per most of the fire safety studies. According to research, building occupants can only safely flee when visibility is at least 4 m (Gager III and Dominguez, 2016). It can be seen from Figs. 5 & 6 that the visibility is lower at the place of fire i.e., at the center. The visibility at human height gets reduced to less than 4 m in about 160 seconds. The main cause of the reduction in visibility is the smoke which initially accumulates near the ceiling and then starts to spread and drops near to human height. Therefore, occupants need to be trained for emergencies to not run haphazardly to avoid stampede situations. Also, illuminated fire safety signage must be installed at the escape routes.

3.2.3. CO₂ and CO concentration

Inhalation of toxic gases released during fire resulted in more deaths compared to any other hazard. The toxic gases can be carbon dioxide (CO₂), carbon monoxide (CO), hydrogen cyanide (HCN), hydrogen chloride (HCl), nitrogen dioxide (NO₂), etc. depending upon the combustibles. The most common among them are CO₂ & CO. The concentration of CO tends to be higher in underventilated conditions because of the partial oxidation of combustibles. The current permissible exposure limit (PEL) recommended by the Occupational Safety and Health Administration (OSHA) for CO₂ and CO are 5000 ppm time-weighted average (TWA) and 50 ppm TWA respectively. The TWA is calculated based on the 8-hour work shift of an employee for a 40-hour work week. The Immediately Dangerous to Life or Health Concentrations (IDLH) set by the National Institute for Occupational Safety and Health (NIOSH) are 40,000 ppm and 1200 ppm respectively. It is noted that the IDLH of both gases is reached only at the center at ceiling height. However, occupants have to escape fast out of the building to prevent continued exposure and thus asphyxiation. Another limit set by NIOSH is the Short-term exposure limit (STEL) which is for 15 minutes’ exposure limit. The STEL should not exceed 30,000 ppm and 200 ppm for CO₂ and CO respectively. This exposure could cause nausea and headache (National Institute for Occupational Safety & Health (NIOSH), 2007). STEL was considered to be the tenability limit. CO₂ and CO concentration reached the tenable limits at human height in 383 and 176 s respectively (Figs. 5 & 6).

3.2.4. ASET and FED calculations

The ASET was estimated from the fire hazards computed in section 3.2. ASET was considered as the least time in which any of the fire hazards would reach to tenable limit (untenability). The tenability criteria for safe evacuation are given in Table 4. In this case, temperature, visibility, CO_2, and CO concentration reached the tenable limit in approximately 170, 160, 383, and 176 s as shown in Fig. 6. The lowest time for untenability among these was of visibility i.e., 160 s. Thus the occupants have to escape out in this time frame only to avoid incapacitation due to fire hazards.

Table 4

Tenability Criteria for Safe Evacuation
Fire hazard	Limit	Reference
Temperature (°C)	< 60°C air temp. (to protect respiratory tract burning)	Thunderhead Engineering Consultants (Gager III and Dominguez, 2016)
Visibility (m)	> 4 m (for familiar small enclosure)	Thunderhead Engineering Consultants (Gager III and Dominguez, 2016)
CO₂ conc. (ppm)	< 30,000 ppm (STEL)	NIOSH (National Institute for Occupational Safety & Health (NIOSH), 2007)
CO conc. (ppm)	< 200 ppm (STEL)	NIOSH (National Institute for Occupational Safety & Health (NIOSH), 2007)

ASET can also be considered as the time at which any of the fractional effective dose (FED) due to radiative and convective flux, fractional incapacitating dose (F_IN) of airborne toxic contaminants such as asphyxiating gases (CO₂ and CO) and fractional effective concentration (FEC) due to smoke obscuration reaches the value above 1. ASET should always be greater than RSET for the tenability condition. FED, F_IN and FEC representing thermal, toxic and visibility hazards were calculated using equations (5)–(13) (O’Connor, 2019). Table 5 shows the input simulated parameters obtained from detectors in the temple and the result of hazard analysis. The thermal and visibility hazards due to radiative and convective heat flux and smoke obscuration reached above 1 in 2 min only. Therefore, on the basis of FED calculation, ASET was between 1 and 2 minutes i.e., approximately 1 minute earlier than graphical prediction. The difference in ASET was attributed to the different threshold limits considered in graphical and FED results. However, the trend of reaching the untenability was similar as visibility and thermal hazards reached untenability before toxic hazards in both methods.

Table 5

Fire Hazard Calculations for the First 5 Minutes of Fire
Input parameters	Time (min)
Input parameters	1	2	3	4	5
Temp (°C)	20.22	23.54	44.45	79.72	96.99
Heat flux (kW/m²)	0.20	1.71	5.52	8.27	8.71
CO (ppm)	19.27	0.92	61.47	143.23	261.07
CO₂ (%)	0.03	0.04	0.20	0.58	1.05
O₂ (%)	20.78	20.78	20.66	20.12	19.51
Smoke (OD/m)	0.18	0.84	0.99	0.99	0.99
t_rad	86.36	4.89	1.03	0.60	0.56
t_conv	1816.25	1082.98	124.73	17.11	8.79
FED_heat	0.06	1.02	4.89	8.59	9.46
FI_CO	0.001	0.000	0.002	0.005	0.009
FI_O2	0.000	0.000	0.000	0.000	0.001
V_CO2	1.049	1.050	1.082	1.163	1.273
F_IN or	0.001	0.000	0.002	0.006	0.012
FI_CO2	0.002	0.002	0.002	0.003	0.004
FEC_smoke	0.945	4.245	4.955	4.99	4.995

3.3. Effect of soot yield and ventilation conditions on fire hazards

The amount of soot created per unit mass of fuel burned is referred to as soot yield. The effect of soot yield and ventilation sensitivity was carried out by changing them within acceptable bounds and leaving other boundary conditions untouched to examine how variations in them impact the fire hazards. The tenability limits are shown with dashed lines in the graphs of fire hazards in Figs. 7 & 8. The range of soot yield (g/g_fuel) as shown in Table 2 was selected based on the following studies (Hou, 2011; Węgrzyński and Vigne, 2017). The highest value of soot yield was taken to be 0.10 g/g_fuel as the value above it barely affects the results. It can be seen from Fig. 7 that changing the soot yield has affected mostly the visibility. The visibility was reaching untenability faster in case of high soot yield. Earlier research has also proved that soot yield has the largest impact on visibility (Węgrzyński and Vigne, 2017). Soot in smoke limits visibility by absorbing light and scattering it in various directions. The higher the soot output, the more light is absorbed and the visibility decreases.

The presence of ventilation in any building allows the exchange of contaminated air with fresh air. The toxic contaminants generated during a fire could diffuse easily in the presence of more ventilation, thus lowering the fire hazards. An underventilated fire is one in which the spread of the fire gets slowed due to lack of oxygen and an accumulation of highly combustible gases gets formed within a confined fire area. This type of fire is more dangerous than a well-ventilated fire because it can lead to poisonous gas accumulation and the potential for flashover. The flashover occurs at a temperature when the upper portion of the room or compartment reaches 600°C. The effect of ventilation (here the number of doorways) on the fire hazards of the temple is represented in Fig. 8. The time to reach untenability in all the fire hazards was lower in the case of one doorway. It can be seen the increment in the number of doorways helped in lowering the vulnerability to all the hazards. A temperature of around 740°C was reached in one doorway opening showing the occurrence of flashover.

3.4. Egress model

RSET depends on various factors such as fire detection time, number of occupants allowed inside the hall and the premises of the temple, occupants' behavior, and wayfinding, which was computed using egress modeling. At the holy places of worship especially during festivals, the maximum limit of occupants is sometimes ignored because of the spiritually motivated crowd. In such cases, the risk of falling short of ASET may occur (Subramanian and Verma, 2022). Simultaneously, there is a huge risk of stampede which further increases the chance of fatalities (Vanumu, Laxmikant and Rao, 2020). Therefore, crowd control is another very important aspect, however, it is not the scope of the present study.

The occupant load factor of 1.4 m²/person was selected according to less concentrated assembly without seats (Bureau of Indian Standards, 2016) with an occupant’s average movement speed of 1.2 m/s (O’Connor, 2019). The floor area of approximately 230 m² (19.25 m $\times$ 19.25 m minus corner areas) can accommodate 165 occupants for the selected occupant load factor. The pre-movement time of 40 s was assumed based on fire safety experience and earlier study (Vistnes, Grubits and He, 2005). The movement time with 2 doorway openings, estimated with the Pathfinder model was found to be 69 s. Therefore, the total egress time for safe evacuation i.e., RSET was 109 s with a safety margin of 51 s. The evacuation of occupants is shown in Fig. 9 at different times. For the selected occupant load factor, the computed RSET was less than ASET which did not affect the occupants' evacuation. However, in case of a high occupant load factor (for example during festivals), all the occupants may not be able to complete the safe evacuation due to incapacitation from the fire hazards. In the worst, possible case, deaths could occur. Therefore, it is necessary to keep a check on occupants' entry limits inside the building or else use active/ passive measures for enhanced safety.

3.5. Smoke controlled by smoke extractor

Smoke is a complex mixture of particulates, gases, and vapors created by the combustion of materials. A survey conducted by NFPA in 2011 finds that the fatalities because of smoke inhalation with respect to burns were in the ratio of 2:1 (National Fire Protection Association, 2015). Temperatures of 1100 degrees Fahrenheit or more can be reached by a fully formed indoor fire. This air is so heated that even one breath might be fatal. No matter if the gases are poisonous, breathing in superheated gases can burn the lungs. Therefore, smoke control is pivotal in saving the lives of people.

Here the application of two smoke extractors was examined in order to increase the ASET. These were located on the side walls opposite to each other at 2 feet below ceiling height with a flow rate of 10 m³/s. The smoke obscuration rate for activation was kept at 3.28% /m. Smoke extractors helped remove the smoke but with increased dispersion of smoke. The high dispersion of smoke, however, did not raise concerns about its potential impact on occupant evacuation due to a decrease in smoke obscuration and thus fire hazards (Fig. 10). The time to reach the tenability limit for fire hazards with two smoke extractors is shown in Fig. 10. It can be seen no fire hazard reached the tenability limit in 600 s at human height. Smoke extractors helped in the reduction of the accumulation of smoke inside the temple and thus the formation of hot layers below the ceiling (Fig. 11). Thereby, reducing the severity of all the fire hazards and helping in increasing the ASET greatly.

FDS proved to be an effective performance-based tool in analyzing the fire safety situation of the temples and finding the untenable conditions. The present study identified the sensitivity of fire growth coefficient, soot yield, and ventilation on fire hazards viz. temperature, visibility, and concentration of toxic CO₂ and CO. The time to reach untenability for each fire hazard was predicted. The ASET for the ultrafast fire was found to be 160 s and the RSET for the occupant load factor of 1.4 m²/person was found to be 109 s. However, RSET could be higher during festivals due to high occupant load. Therefore, for the life safety of occupants, it was suggested to use smoke extractors wherever possible and set the entry limit of the occupants to avoid incapacitation and stampede. The use of two smoke extractors was shown to decrease the severity of the fire hazards, enhancing the tenability and ASET, thus increasing the safety margin between ASET and RSET greatly. A simple methodology for improving the tenability condition in heritage buildings was proposed as shown in Fig. 12. We expect this study could serve as a basis for the precautionary measures to be taken in the heritage temples and other heritage structures for the protection of the structure as well as the occupants from the fire.

Ethical approval Not applicable.

Consent to participate Not applicable.

Consent for publication The authors have given their consent to publish the manuscript.

Author contributions

Aravind Kumar: study's conception, supervision, review, and editing. Mohammad Ahmad Shoeb: study's conception, simulations, analysis and initial draft preparation. Mahesh Kumar Tiwari: supervision and review. Rakesh Kumar: review and editing. Charu Mehta: review and editing. All authors read and approved the final manuscript.

Funding

This research is framed within the Grant-in-Aid Project (GAP-0311), sponsored by the Department of Science and Technology, India. The authors would like to acknowledge the Council of Scientific & Industrial Research Institute, Delhi, and the Director, Central Building Research Institute, Roorkee, India for financial support and permission for the activity.

Competing interest

No potential conflict of interest was reported by the authors.

Ball, P. (2020) ‘The huge scientific effort to study Notre-Dame’s ashes’, Nature, 577(7789), pp. 153–154. doi:10.1038/d41586-020-00008-5.
Bernardini, G. (2017) Chapter 2 - Fire Safety of Historical Buildings: The Occupants Perspective, Fire Safety of Historical Buildings. SpringerBriefs in Applied Sciences and Technology (SpringerBriefs in Applied Sciences and Technology). doi:10.1007/978-3-319-55744-1.
Betting, B. et al. (2019) ‘Experimental and numerical studies of smoke dynamics in a compartment fire’, Fire Safety Journal, 108(May), p. 102855. doi:10.1016/j.firesaf.2019.102855.
Bukowski, R.W. and Nuzzolese, V. (2009) ‘Performance-based Fire Protection of Historical Structures’, Fire Technology, 45(1), pp. 23–42. doi:10.1007/s10694-008-0044-x.
Bureau of Indian Standards (2016) ‘National Building Code Of India , Part IV, Fire and Life Safety’. New Delhi: Bureau of Indian Standards (BIS).
Donghyun, K. et al. (2020) ‘A Study on Fire Vulnerability Analysis of Wooden Cultural Property Using Fire Simulation – For Bulguksa Temple (Korean)’, Journal of the Society of Cultural Heritage Disaster Prevention, 5(3), pp. 233–242. Available at: https://www.kci.go.kr/kciportal/ci/sereArticleSearch/ciSereArtiView.kci?sereArticleSearchBean.artiId=ART002672350.
Fire risk in heritage properties (2021) Ecclesiastical Insurance Office plc (EIO). Available at: https://ecclesiastical.com/documents/heritage-fire-introduction.pdf (Accessed: 17 February 2015).
Gager III, A.H. and Dominguez, G. (2016) ‘Tenability criteria in unique situations and atypical buildings’, Fire and Evacuation Modelling Technical Conference, pp. 1–14. Available at: https://www.thunderheadeng.com/files/com/FEMTC2016/files/d1-14-gager/gager-paper.pdf.
Garcia-Castillo, E., Paya-Zaforteza, I. and Hospitaler, A. (2021) ‘Analysis of the fire resistance of timber jack arch flooring systems used in historical buildings’, Engineering Structures, 243, p. 112679. doi:10.1016/j.engstruct.2021.112679.
Govindrajan, V. (2018) ‘Fire in Madurai’s Meenakshi temple sparks demand for shops to be evicted from complex’, Scroll India, 6 February. Available at: https://scroll.in/article/867574/fire-in-madurais-meenakshi-temple-sparks-demand-for-shops-to-be-evicted-from-complex.
Van Hees, P. (2013) ‘Validation and verification of fire models for fire safety engineering’, Procedia Engineering, 62, pp. 154–168. doi:10.1016/j.proeng.2013.08.052.
Hou, J.Y. (2011) Distribution Curves for Interior Furnishings on CO2, CO, HCN, Soot and Heat of Combustion. University of Canterbury Christchurch,. Available at: https://ir.canterbury.ac.nz/server/api/core/bitstreams/63ca3eb7-4c6d-463a-a9d6-326ce9c8768d/content.
Huai, C. et al. (2021) ‘Experimental and Numerical Analysis of Fire Risk in Historic Chinese Temples: A Case in Beijing’, International Journal of Architectural Heritage, pp. 1–15. doi:10.1080/15583058.2021.1916648.
Jiji, K. (2020) ‘Fire at Okinawa’s Shuri Castle barely damaged original ruins’, The Japan Times, 31 January. Available at: https://www.japantimes.co.jp/news/2020/01/31/national/shuri-castle-fire-minor-impact-universal-value-cultural-affairs-agency/.
Khan, A.A., Usmani, A. and Torero, J.L. (2021) ‘Evolution of fire models for estimating structural fire-resistance’, Fire Safety Journal, 124, p. 103367. doi:10.1016/j.firesaf.2021.103367.
Kumar, A. et al. (2022) ‘Experimental and numerical simulation studies of liquefied petroleum gas fire in a full–scale compartment’, Process Safety Progress, 41(1), pp. 195–206. doi:10.1002/prs.12292.
Lim, J.W. et al. (2019) ‘Numerical analysis of performances of passive fire protections in processing facilities’, Journal of Loss Prevention in the Process Industries, 62(July), p. 103970. doi:10.1016/j.jlp.2019.103970.
Liu, Q. et al. (2015) ‘Numerical studies on kitchen fire hazards with multiple burning sources’, Building Simulation, 8(4), pp. 453–463. doi:10.1007/s12273-015-0223-x.
Long, X., Zhang, X. and Lou, B. (2017) ‘Numerical simulation of dormitory building fire and personnel escape based on Pyrosim and Pathfinder’, Journal of the Chinese Institute of Engineers, Transactions of the Chinese Institute of Engineers,Series A, 40(3), pp. 257–266. doi:10.1080/02533839.2017.1300072.
Marrion, C.E. (2016) ‘More effectively addressing fire/disaster challenges to protect our cultural heritage’, Journal of Cultural Heritage, 20, pp. 746–749. doi:10.1016/j.culher.2016.03.013.
McGrattan, K. et al. (2013) Fire Dynamics Simulator Technical Reference Guide Volume 1: Mathematical Model, National Institute of Standards and Technology. Available at: https://nvlpubs.nist.gov/nistpubs/SpecialPublications/NIST.SP.1018e6.pdf.
McGrattan, K et al. (2013a) Fire Dynamics Simulator Technical Reference Guide Volume 2: Verification, National Institute of Standards and Technology.
McGrattan, K et al. (2013b) Fire Dynamics Simulator Technical Reference Guide Volume 3: Validation, National Institute of Standards and Technology.
McGrattan, K. et al. (2013) Fire Dynamics Simulator User’s Guide. Special Publication (NIST SP) - 1019. Available at: https://ws680.nist.gov/publication/get_pdf.cfm?pub_id=913619.
Mega, E.R. (2019) ‘After the flames’, Nature, 571, pp. 312–315. Available at: https://www.nature.com/articles/d41586-019-02141-2.pdf.
Morgan J. Hurley (ed.) (2016) SFPE Handbook of Fire Protection Engineering. Society of Fire Protection Engineering. Available at: https://www.springer.com/gp/book/9781493925643 (Accessed: 13 October 2015).
Natampally, M. (2018) Making of Hampi—An Attempt to Bridge Culture and Technology, Digital Hampi: Preserving Indian Cultural Heritage. Springer Singapore. doi:10.1007/978-981-10-5738-0.
National Fire Protection Association (2015) NFPA 92: Standard for Smoke Control Systems.
National Fire Protection Association (2019) NFPA 72: National Fire Alarm and Signaling Code.
National Institute for Occupational Safety & Health (NIOSH) (2007) ‘NIOSH Pocket Guide To Chemical Hazards’. Cincinnati. Available at: https://www.cdc.gov/niosh/docs/2005-149/pdfs/2005-149.pdf.
O’Connor, D.J. (2019) SFPE Guide to Human Behavior in Fire, SFPE Guide to Human Behavior in Fire. Society of Fire Protection Engineering. doi:10.1007/978-3-319-94697-9.
Revenue and Disaster Management (2016) ‘Memorandum-Puttingal temple fire cracker explosion’. Kollam, India: Government of Kerela, India. Available at: https://sdma.kerala.gov.in/wp-content/uploads/2018/11/1.Memorandum-Puttingal-Temple-2016.pdf.
Su, H.-C. et al. (2020) ‘Survey and Experimental Investigation of Movable Fire Loads in Japanese-Style Wooden Historical Buildings’, International Journal of Architectural Heritage, 14(6), pp. 931–942. doi:10.1080/15583058.2019.1587040.
Subramanian, G.H. and Verma, A. (2022) ‘Crowd risk prediction in a spiritually motivated crowd’, Safety Science, 155. Available at: https://doi.org/10.1016/j.ssci.2022.105877.
The Hindu (2021) ‘Mulamkadakam temple suffers damage in fire’, 24 January. Available at: https://www.thehindu.com/news/national/kerala/mulamkadakam-templesuffers-damage-in-fire/article33645961.ece.
Torero, J.L. (2019) ‘Fire Safety of Historical Buildings: Principles and Methodological Approach’, International Journal of Architectural Heritage, 13(7), pp. 926–940. doi:10.1080/15583058.2019.1612484.
Vanumu, L.D., Laxmikant, K. and Rao, K.R. (2020) ‘Human stampedes at mass gatherings: An overview’, Collective Dynamics, 5, p. A83. doi:10.17815/CD.2020.83.
Vistnes, J., Grubits, S. and He, Y. (2005) ‘A Stochastic Approach To Occupant Pre-movement In Fires’, Fire Safety Science, 8, pp. 531–542. doi:10.3801/IAFSS.FSS.8-531.
Węgrzyński, W. and Vigne, G. (2017) ‘Experimental and numerical evaluation of the influence of the soot yield on the visibility in smoke in CFD analysis’, Fire Safety Journal, 91(February), pp. 389–398. doi:10.1016/j.firesaf.2017.03.053.
Wen, J.X. et al. (2007) ‘Validation of FDS for the prediction of medium-scale pool fires’, Fire Safety Journal, 42(2), pp. 127–138. doi:10.1016/j.firesaf.2006.08.007.

Fire safety in heritage structures: A case study on heritage temple with performance-based design fire approach

Status:

Version 1

Abstract

Figures

1. Introduction

2. Modeling and simulation

2.1. Details of the temple geometry and boundary conditions

2.2. Description of fire scenario

2.3. Methodology

3. Results and discussion

3.1. Grid sensitivity analysis

3.2. Fire and smoke development

3.2.1. Temperature rise at different locations

3.2.2. Visibility

3.2.3. CO₂ and CO concentration

3.2.4. ASET and FED calculations

3.3. Effect of soot yield and ventilation conditions on fire hazards

3.4. Egress model

3.5. Smoke controlled by smoke extractor

4. Conclusions

Nomenclature

Declarations

References

Status:

Version 1

Model quantities	Description
Model size	19.25 m \(\times\) 19.25 m \(\times\) 9.75 m
Size of the fire source	4 m² (2 m \(\times\) 2 m)
Vent size	1.5 m \(\times\) 3.5 m approx.
Number of vents	1–3 doorway
Size of mesh in rest of the model	0.25 m
Size of mesh around fire source	0.0625–0.125 m
Heat of combustion	20 MJ/kg
Radiative fraction	0.35
HRRPUA	1.25 MW/m²
Max HRR	5 MW
Fire growth coefficient (α)	0.012 (medium) – 0.19 (ultrafast) kW/s²
Soot yield	0.02–0.10 g/g_fuel
CO yield	0.05 g/g_fuel
Simulation time	600 s
Simulation type	LES
Ambient temperature	20°C
Ambient pressure	1 atm

Fire safety in heritage structures: A case study on heritage temple with performance-based design fire approach

Status:

Version 1

Abstract

Figures

1. Introduction

2. Modeling and simulation

2.1. Details of the temple geometry and boundary conditions

2.2. Description of fire scenario

2.3. Methodology

3. Results and discussion

3.1. Grid sensitivity analysis

3.2. Fire and smoke development

3.2.1. Temperature rise at different locations

3.2.2. Visibility

3.2.3. CO2 and CO concentration

3.2.4. ASET and FED calculations

3.3. Effect of soot yield and ventilation conditions on fire hazards

3.4. Egress model

3.5. Smoke controlled by smoke extractor

4. Conclusions

Nomenclature

Declarations

References

Status:

Version 1

3.2.3. CO₂ and CO concentration