Investigation of the effects of a car fire in a basement car park using computational fluid dynamics analysis
Ang, Wei Jian
Date of Issue2018
School of Mechanical and Aerospace Engineering
Generally, all building developments are required to fully comply to the stipulated codes and regulations set by authorities like Singapore Civil Defence Force (SCDF) and Building & Construction Authority (BCA). For building developments that does not comply to the stipulated codes and regulations, a performance-based study is required to be conducted to show the authorities that it does not affect the life safety of the people who are evacuating from a fire scene and the fire fighters entering the scene to fight the fire. This report discussed the formulation and the simulation methodology of seven fire scenarios located at different levels of the basement car park development using computational fluid dynamics (CFD) solver software. Results of the CFD simulation were used to estimate the Available Safe Egress Time (ASET) for the occupants to escape from the development. To find the Required Safe Egress Time (RSET) for the occupants to escape the development, calculation of the longest travel time from the most remote point of the development to the nearest escape staircase; time taken to queue to exit through the escape staircase; fire detection time; occupants’ reaction time were taken into consideration. SCDF required a safety factor of more than two for the ratio of the ASET and RSET (i.e. ASET ÷ RSET > 2); lower average temperature (i.e. from finished floor to 2.5m height) to be below 60℃; upper average temperature (i.e. from 2.5m height to ceiling) to be below 100℃ and visibility of more than 10m at 2.5m above finished floor level as the acceptance criteria for performance-based study. All seven fire scenarios discussed in this report met the criteria. To validate the simulation conducted, a hot smoke test was conducted for one of the simulated fire scenario (the rest of the basement car park levels are still under construction at point of writing this report) witnessed by an authorised representative from SCDF. The hot smoke test did not match with the simulated results obtained with probable explanation like not meeting the minimum required perforation of the false ceiling panel; additional pipes, ducts and plenum boxes installed in the level that was not provided during the CFD modelling and before the simulation phase; the heat release rate of the simulation and the actual hot smoke test affecting the buoyancy of the smoke. The representative expressed his concern if the designed performance based smoke purging system can support and meet the requirements of the prescribed smoke control system in an event of a fire. Considering the cost and time spent in the installation of these items and the targeted operation date, redesigning the duct layout, or changing of the perforated false ceiling will have negative impact on the progress of the development. Hence, the new objective would be to improve the smoke extraction and minimise the negative impact to the progress of the development. The proposed alternative solution involves removing the perforated false ceiling panels directly below the supply and exhaust air grilles and increasing the number of extraction points. Another CFD study was conducted to access the feasibility of the proposed alternative solution. Simulation results showed that it met the acceptance criteria set by SCDF for performance-based study. To explore the effects of the temperature and visibility of the car park if the emergency mode is unable to activate, a CFD study was conducted to find out the effects. One of the fire scenario was used to conduct this study and the comparison of the results obtained showed that the higher the air change rate of air within the level, the lower the average temperature of the computational domain. With the discrepancy of the hot smoke test and the simulated results, a literature study was conducted on a conference paper. This conference paper documented a lab-based experiment by burning various fuels with different soot yield and measurement of the visibility and temperature. In addition, CFD simulation was conducted using various solver software in the market. Generally, it was observed that the CFD solver software overestimated the temperature and underestimated the visibility when compared to the measured results of the lab test. Hence, it was inferred that the CFD results are slightly more onerous than the actual scenario in the event a fire.
Final Year Project (FYP)
Nanyang Technological University