EXPERIMENTAL AND NUMERICAL STUDIES ON COMPARTMENT FIRES USING DIESEL POOL FIRES AT VARIOUS ELEVATIONS AND LOCATIONS

Abstract

The present research work tries to study behaviour of fire inside a built environment, both experimentally and numerically in a naturally ventilated compartment. The test compartment is a size of 4 m length, 4 m width, 4 m height with a door 2 m height (Hd) and 1 m width (Wd). In total thirty-seven experiments were conducted using diesel fuel with different pan diameters ranges from 0.2 m to 1 m. The fire source elevation (h) was the principal parameter investigated in the present work. In order to study the behaviour of various fuel position, a large pool diameter 0.8 m was burned in centre, corner and rear wall centre along with six different fire source elevation (i.e. h/Hd = 15%, h/Hd = 30%, h/Hd = 45%, h/Hd = 60%, h/Hd = 75% and h/Hd = 90%). During experimentation, it was observed that variation in pan height significantly affects the mass loss rate (MLR), heat release rate (HRR), flame height, species production, compartment temperature, smoke layer height, plume residence time, incident heat flux, doorway temperature and velocity. For the same fuel, the experimental mass loss flux (mass loss rate per unit pool surface area) was found to be 0.23 times less than the mass loss flux obtained with open fire test. A correlation was developed to estimate the flame height for a compartment fire as the function of pool diameter and heat release rate. Heat flux variations were observed to follow power law trends for all inner surfaces of the compartment with respect to pool diameters. The effect of fire source elevation on burning rate of pool fires were different according to diameter of pool fires. Pool diameters of 0.2 m and 0.4 m were not shown significant variation in burning rate of pool fire along the fuel pan elevation. For a pool diameter of 0.8 m elevated at higher elevation (i.e. h/Hd = 90%), the MLR decreased drastically, to half of the maximum MLR value obtained in case of lower elevation (i.e. h/Hd = 15%).With respect to the developed thermal environment, the fire scenario inside the compartment seems to be less hazardous in case of higher elevated fire. Results of CO2/CO ratio shows exponential decay trend with fuel pan elevation, the maximum concentration of CO increased with increase in ceiling temperature and showed a parabolic trend. Moreover, lower CO2/CO ratio implies incomplete combustion and thereby leads to higher toxicity inside the compartment. Large-scale pool fire (D = 0.8 m) located at center gives more HRR value for fire as compared to those located at corner or rear wall center, at lowest elevated position. Further increase in pan height, the HRR is not significantly changed with respect to location up to critical dimensionless height, h/Hd = 60%. The maximum value of ceiling heat flux varied from 13 to 18 kW/m2 for h/Hd = 30% to 60%, which crosses the threshold value of 12.6 kW/m2 that may be cause of wood ignition after a long exposure. With increasing pan height placed at center, the temperature first goes to peak and then decreases. The highest temperature registered was at h/Hd = 60%, 75% and 90% for 0.8 m, 0.6 m and 0.4 m pool diameter respectively. However, the temperatures were found to be almost same at all the elevations of 0.2 m pool diameter. The radiative heat flux by the hot gas was found to be higher for the compartment fire having less oxygen available at the flame base. A validation of the point source model through measured heat flux values on walls, 1 m above the floor, has shown good agreement for large size pool fire while deviation is found to be more in case of small size of pool fires. Hence, the percentage of hot gas heat flux increases with increasing pan height. An attempt has also been made to validate the experimental results of centrally located 0.8 m pool fire by making an energy balance for h/Hd = 15% at which fuel being burnt completely by resulting maximum heat release rate. Various heat loss terms (boundary heat loss, doorway heat loss and gas heating) are calculated to predict the heat release rate. A major portion of heat release goes to heat up the compartment boundaries, approximately one fourth of energy has been lost through doorway opening and the remaining has been used for gas heating. Numerical simulations were performed using Fire Dynamic Simulator (FDS), version 6.7.1 developed by NIST, USA. Heat release rate measured by oxygen depletion calorimeter is specified as the input in FDS. Simulations were performed with radiation fraction 0.25 as reported earlier for diesel fuel. The value of 𝐷∗𝑑𝑥⁄ would be suggested 13 for accurately resolving the events of compartment fire. The major contribution of the study is the compartment size, which is much larger than the standard international standard organization (ISO) room. Such experimental data is especially useful with modern construction trends of high ceiling large floor plans. The present study will be helpful for designing fire safety methods and consequently for developing fire safety standards. Moreover, results will be useful for optimizing the fire protection systems such as detectors location, location and size of extracting ports and fans etc. Also, evaluation aforesaid will reduce the impact of fire on people, property and environment. Keywords: Compartment fire, pool fire elevation, diesel fuel, large scale heat release calorimeter, flame height, heat flux, global equivalence ratio, CO2/CO ratio, plume residence time, fractional effective dose, fire dynamic simulator.