Studies On Cabin Fire And Smoke Under Mechanical Exhaust And Sprinkler In Large Space Buildings

Combustibles in many modern large space buildings are not stored evenly on the ground due to functional requirements. They are stored in small spaces, which would give some areas with high fire load. An obvious example is using the atrium level as retail shops and small restaurants. A new fire protection design concept on cabins appeared in the 90s. Area with high fire load in a large space building is isolated from other parts of the building as a cabin. Effective fire suppression system would be put in a cabin. Fire and smoke are prevented to spill out to the large space for ensuring total fire safety.It is a challenging problem in designing fire safety projects with a cabin. Fire protection systems include the fire detection and alarm system, sprinkler system, and mechanical smoke exhaust system. Fire detection and alarm system is used for detecting the fire promptly. Sprinkler system is used to suppress the fire, and control the fire size if not able to extinguish it. Mechanical smoke exhaust system should be able to extract smoke to keep the smoke layer higher than the height of the ventilation opening, so that there is no spilling of smoke to the large space.In this thesis, a cabin with supply air from outside was considered. The smoke production rates in a cabin fire under mechanical smoke exhaust predicted by the plume models by Zukoski, McCaffrey, and Heskestad were analyzed with full-scale burning tests with big fires. Results indicated that the prediction by the Heskestad model fitted the experimental data well. By fitting to the experiment, the Heskestad model was revised preliminarily. A semi-empirical correlation of the fan extraction rate with the heat release rate and smoke layer height for preventing smoke from spilling out from the cabin under a big cabin fire was derived. Further, the efficiency of mechanical smoke exhaust in the cabin was analyzed. The three factors affecting the efficiency of mechanical smoke exhaust on the heat release rate, smoke layer height and the fan extraction rate were studied. Heat release rate is confirmed to be the key factor. An efficiency of 60% for the mechanical smoke exhaust was recommended based on the experiment as a reference data to design mechanical smoke exhaust system in a cabin with supply air from the surrounding.With mechanical smoke exhaust at high level, the predicted smoke production rates based on the McCaffrey model under different smoke layer heights were obtained by revising the coefficient. The revised coefficient changed with the smoke layer height. When the smoke layer height was high, the revised coefficient was higher than 1, at the buoyant plume region. When the smoke layer height was low, the revised coefficient was lower than 1, at the continuous flame region. When the smoke layer height was at other locations, the revised coefficient was higher than 1 first, then decreased with decreasing smoke layer height to below 1. The overall revised coefficient was about 1, at the intermittent flame region. From the experimental data, the revised coefficient was about 1.84 at the buoyant plume region, about 1 at the intermittent flame region, and about 0.67 at the continuous flame region. An empirical exponential correlation of the revised coefficient with the smoke layer height was derived. A semi-empirical correlation of the fan extraction rate with the heat release rate and smoke layer height for preventing smoke from spilling out from the cabin with mechanical smoke exhaust at high level was deduced.With mechanical smoke exhaust at low level, the smoke layer interface was low, falling almost to the continuous flame region. The efficiency of mechanical smoke exhaust decreased greatly because more air was extracted directly from the lower air layer by the fan. The efficiency of the mechanical smoke exhaust increased with increasing heat release rate. An empirical exponential correlation on the efficiency of mechanical smoke exhaust with the heat release rate was derived. From the experimental data, the average efficiency of mechanical exhaust was about 26.5%. The model for predicting mechanical smoke exhaust rate was obtained based on the McCaffrey model. A semi-empirical correlation of the fan extraction rate with the heat release rate and smoke layer height for preventing smoke from spilling out from the cabin with mechanical exhaust at low level was also deduced. Under the same conditions, the flow rate of mechanical smoke exhaust at low level was about 5 times than that at high level. The computational fluid dynamics software Fire Dynamics Simulator (FDS) was used to simulate cabin fire with mechanical smoke exhaust at low level. Though the calculated average smoke layer temperature was higher than the experimental data, the predicted flow field can give insight on how much air was extracted. The predicted average smoke layer heights in the cabin with mechanical smoke exhaust at low level fitted the experimental data well.Full-scale burning tests were carried out to study cabin fire and smoke under the operation of sprinkler. A critical pressure was found on controlling or extinguishing the fire under the operation of sprinkler. The critical pressure increased with increasing heat release rate. The average smoke temperature difference, O₂ concentration, CO concentration in the cabin and the pattern of smoke spilling out from the cabin and then filling the large space under different pressures were obtained. Results showed that the average smoke temperature difference in the cabin would decrease greatly and the fire was suppressed. The O₂ concentration first decreased to the lowest concentration when the sprinkler was operated, then increased. The O₂ concentration increased suddenly when extinguishing the fire, but increased continuously when controlling the fire. After the sprinkler was activated, the CO concentration increased sharply. Two peaks were found when controlling the fire, only one peak was observed when extinguishing the fire. When the sprinkler pressure was lower than the critical pressure, the CO production rate increased with increasing sprinkler pressure. When the sprinkler pressure was higher than the critical pressure, the CO production rate decreased with increasing sprinkler pressure. The CO production rate would reach its maximum at the critical pressure. The average smoke temperature difference in large space buildings was very low, but smoke spilled out from the cabin could fill the large space buildings quickly after the sprinkler was activated. Evacuation would be affected. The trend of smoke production rate was similar to the CO production rate.Full-scale burning tests were also carried out to study cabin fire and smoke under the operation of sprinkler and mechanical smoke exhaust. Different combinations of the pressure and fan extraction rate were found on controlling or extinguishing the fire. The average smoke temperature difference, CO concentration in the cabin and the smoke filling pattern in the large space were obtained. The pressure of sprinkler was the key factor, and the fan extraction rate was the secondary factor. The effects of these two factors were contrary. The concept of equivalent pressure was presented to replace the contrary effect of the fan extraction rate. So the trends of cabin fire and smoke under the operation of sprinkler and mechanical smoke exhaust would be consistent with that under the action of sprinkler due to equivalent pressure. The effect of fire control was validated with full-scale burning tests. Results showed that the CO and smoke production rate increased with increasing sprinkler pressure, and decreased with increasing extraction rate of the fan. The effect of extinguishing the fire was deduced. The CO and smoke production rate decreased with increasing sprinkler pressure, and increased with increasing extraction rate of the fan. The CO and smoke production rate would reach their maximum at the critical combinations of the pressure and extraction rate of the fan.