Performance Optimisation of Water Mist Fire Suppression System Design

Description

With the merits of being environmentally friendly, inexpensive and effective in fire suppression, and non-toxic in nature, the water mist fire suppression system (WMFSS) is one of the possible options for replacing halon-based systems. Many potential applications of WMFSS have been identified including fire suppression systems in marine vessels, train cars, atria, museums, and historical buildings where the risk of water damage is a concern. However, only limited information on the effectiveness in relation to size distribution of water mist droplets, nozzle operation, and design parameters is available. In this study, under a wide range of selected practical nozzle operation and design parameters, experiments of water mist spray on fire will be performed in a burning facility instrumented for gases and mass-loss monitoring. The size distribution, velocity distribution, and water mist flux density will be measured by laser Doppler velocimetry/adaptive phase Doppler velocimetry. Based on the experimental results and analysis, a general regression neural network and fuzzy ART novel hybrid-model will be used to develop an objective function. Subsequently, the model using classical generic algorithms as a tool will be developed for the optimization of WMFSS designs. The optimization model, in the long term, can be used to establish the guidelines for WMFSS designs and to facilitate performance based fire engineering practices.