Numerical and experimental study of fire-extinguishing performance of ultra-fine water mist

Abstract

In the past, gas-based fire extinguishing systems (GFES) (halon and carbon dioxide, etc.) have been widely used to protect spaces housing electrical equipment. However, these systems have some deficiencies: false actuation of GFES in an enclosed space may threaten the safety of occupants since fire-fighting implies evacuation. Many studies have focused on extinguishing efficiency of water mist in such spaces. The conventional water mist system has limitations: 1) Difficulty of extinguishing small fires; 2) Difficulty of extinguishing fires in shielded or obstructed spaces; et al. These limitations are mainly associated with high fallout rates of droplets, this tend to significantly decrease the mist concentration especially in regions away from the nozzle spray patterns. Therefore, the present study focus on the fire-extinguishing performance of UFM (UFM) generated by ultra-sonic atomization, the method of improving its fire-extinguishing performance and the flow behaviour of UFM. The work and contributions of this study can be summarised as follows: 1) On minimum extinguishing concentration (MEC) of UFM: The MEC of UFM was modeled based on limiting oxygen concentration (LOC) and combustion limit temperature (CLT), respectively. By analyzing the MEC from the two models, it is concluded that heat absorption is a way with more potential than oxygen dilution in extinguishing fire for UFM. Fire extinguishment experiment was then carried out in a modified cupburner. Tests using the same scenario were repeated many times to record the average and the standard deviation of extinguishing times. For MEC, experimental results agree well with the model based on LOC, and disagree with the model based on CLT, the reason is flow behavior of UFM: a) UFM totally evaporates around the flame and hardly enters flame’s core; and b) only the generated water vapor near flame follows the entraining flow and interacts with the lame. The mist concentration should be higher than a critical value to be able to extinguish the fire with a expected time. However, there is no need for a mist concentration over and above certain threshold because it will not improve the efficiency of the fire extinguishing system. Since a potentially more effective mechanism cannot be activated, fire extinguishing performance of UFM need to be improved by increasing droplets size and adding chemical additives 2) On fire extinguishing performance of UFM system with additives: Based on the distribution of droplets size after ultrasonic atomization and the drag force of a droplet in the airflow, effective mass fraction of UFM was modeled. The model indicated that the effective mass fraction of UFM decreases with the increase of solution surface tension and increase of water temperature can increase fire extinguishing effectiveness of UFM. A simple test was conducted to measure the change of effective mass fraction of UFM. Experimental results agree well with the theoretical analysis, which showed that increase of concentration of metal salt will decreases the mass of UFM, while adding surfactant can increase the mass of UFM. Fire extinguishment experiment was then carried out. The experiments showed the increase of water temperature can increase fire extinguishing performance of UFM system, which agrees with theoretical analysis. Adding a small quantity of a certain saline to the UFM can significantly improve the fire extinguishing performance of UFM system. However, increase of fire extinguishing efficiency by further increase of saline would not be so obvious, since UFM mass is decreased by the increase of surface tension of water solution. Based on the number of ultrasonic atomizer needed, on a mass fraction basis, the order of effectiveness is: K2C2O4> K2CO3> KCl> KHCO3> NaCl> CH3COONa>KH2PO3> Urea. Adding of urea leads to combustion enhancement. A multi- component method is proposed to improve the fire-extinguishing performance of UFM. The method involves adding both a type of metal salt and a surfactant. 3) Numerical study on the interaction diffusion flame with UFM: The empirical model for predicting local extinction based on the oxygen concentration in FDS is unable to determine fire extinguishing efficiency in this study. The EDC-modified model in SIMTEC cannot predict the MEC of UFM. The finite-rate model with detailed chemical reaction (in FLUENT), can simulate flame extinguishment, and predict the MEC of UFM. The primary mechanism of fire extinction in the simulation using FLUENT is oxygen displacement, which agrees with the experiment. Fire intensification was observed in the simulation using FLUENT. The mechanism of fire intensification by NH3 and HNCO is NH3+H=NH2+H2 and NH3+O=NH2+OH; and the primary mechanism of fire intensification by HNCO is HNCO+O=NCO+OH, HNCO+H=NH2+CO and HNCO+OH=NCO+H2O. 4) CFD simulation and experimental study on flow behavior of UFM: The CFD simulations using DPM cannot simulate transportation and flow behavior of low momentum UFM. The dense gas model showed a significant improvement in predicting UFM transportation and flow behavior. Fire of a larger size is easier to extinguish in a compartment space, which can promote transportation of UFM in a compartment. However, a larger size of fire is more difficult to extinguish in a tunnel space, which prevent UFM transportation to the other side of the tunnel in the opposite direction. The effect of obstruction in extinguishing efficiency depends on the location of obstruction. If obstruction is located between the fire source and the mist source, the obstruction would decrease the extinguishing efficiency of UFM. If obstruction is located behind the fire source, the obstruction would improve the extinguishing efficiency of UFM.

Date of Award	3 Oct 2012
Original language	English
Awarding Institution	City University of Hong Kong
Supervisor	Siu Ming LO (Supervisor)