Videos for Wind-Driven Fires: Governors Island & Laboratory Experiments

Governor's Island test building. (Photo credit: NIST)

Together with the Fire Department of New York City (FDNY), the Polytechnic Institute of New York University, and with the support of the Department of Homeland Security Federal Emergency Management Agency Assistance to Firefighters Research and Development Grant Program and the United States Fire Administration, our researchers conducted a series of wind-driven fire experiments in a seven-story building on Governors Island, New York, in February 2008.

The objective of these experiments was to improve the safety of firefighters and building occupants by developing a better understanding of wind-driven fires and wind-driven firefighting tactics, including structural ventilation and suppression. The experiments evaluated the ability of positive pressure ventilation fans (PPV), wind control devices (WCD), and exterior water application via high rise nozzles (HRN), operated from the floor below to mitigate the hazards of a wind-driven fire in a structure.

The building was instrumented to measure temperature, differential pressure, and gas velocity to examine the impact of the PPV fans, WCDs and HRNs. Each of the experiments was documented with video and thermal imaging cameras. These experiments captured video of specific fire phenomena that are not typically observable on the fire ground. These experiments demonstrated the "extreme" thermal conditions that can be generated by a "simple room and contents" fire and how these conditions were extended along a flow path within a real structure when wind and an open vent are present. The thermal conditions generated along the flow path were untenable, even for firefighters in protective gear. Deployed from the floor above the fire, a WCD was shown to be effective in reducing the thermal hazard in the corridor and stairwell. Deployed from the floor below the fire, external water application with a HRN was demonstrated to be effective in reducing the thermal hazard in the corridor and stairwell. FDNY has implemented the use of these tools and tactics throughout the department.

Governors Island: Floor 3, Apartment A (00:56) No audio.

The thirteenth experiment in the structure was located on the 3^rd floor in apartment 3A. Experiment 3A was ignited in both the bedroom and the living room in an effort have both rooms involved in fire to examine the impact of a FBN deployed in the bedroom. This experiment also utilized a simulated wind provided by the MVU. The simulated wind was aimed into the bedroom window with a magnitude of approximately 9 m/s to 11 m/s (20 mph to 25 mph).

Governors Island: Floor 3, Apartment E (00:51) No audio.

The twelfth experiment in the structure was located on the 3rd floor in apartment 3E. Experiment 3E was ignited in both the bedroom and the living room in an effort have both rooms involved in fire simultaneously to examine the impact of a FBN deployed in the bedroom. This experiment also utilized a simulated wind provided by the MVU. The simulated wind was subjected to the bedroom window at a magnitude of approximately 9 m/s to 11 m/s (20 mph to 25 mph).

Governors Island: Floor 3, Apartment G/K (01:30) No audio.

The fourteenth and last experiment in the structure involved two apartments on the 3^rd floor, apartments 3G and 3K. Apartment 3G was on the up wind side of the building and was used as the ignition apartment. The fire was ignited in three locations simultaneously; both of the bedrooms and the living room. In this experiment, the natural wind played a significant role. The wind was coming from the WNW and hitting the side of the building at an average wind speed in the range of 7 m/s to 9 m/s (15 mph to 20 mph), with gusts of up to 13 m/s (29 mph). Later in this experiment a simulated wind provided by the MVU was also utilized. The simulated wind was aimed into the bedroom window with a magnitude of approximately 9 m/s to 11 m/s (20 mph to 25 mph). Apartment K was located on the downwind side of the apartment and was loaded with fuel. The flow path on the 3rd floor was arranged to go from the most remote bedroom in apartment G through the open apartment door into the public corridor into the open door of apartment K to the open bedroom windows on the downwind side of the building.

Governors Island: Floor 7, Apt. G2 (00:31). No audio.

The second experiment in the structure was located in the same apartment as the first experiment, 7G. Experiment 7G2 was ignited in the middle bedroom in a trashcan at the base of the bed, and in the living room on the top of sofa that backed to the kitchen. The fire in the bedroom from the first experiment, Experiment 7G, did not extend beyond the room of origin so the apartment was cleared of smoke and experiment 7G2 was ignited. Both windows in the back bedroom failed in the previous experiment so they remained open during the duration of this experiment. This experiment also utilized natural ventilation conditions with no simulated wind and very little natural wind.

Governors Island: Floor 7, Apt. K (00:58) No audio.

The fifth experiment in the structure was located on the 7th floor in apartment 7K. Experiment 7K was also ignited in the bedroom furthest from the open apartment door. This experiment also utilized a simulated wind provided by the MVU. The simulated wind was subjected to the bedroom double window at a magnitude of approximately 9 m/s to 11 m/s (20 mph to 25 mph).

Eight full scale experiments were conducted to examine the impact of wind on fire spread through a multi-room structure and examine the capabilities of wind control devices (WCD) and externally applied water streams to mitigate the hazard. The principle measurements used to examine the impact of the tactics were heat release rate, temperature, heat flux, and velocity inside the structure. Measurements of oxygen, carbon dioxide, total hydrocarbons and differential pressures were also measured. Each of the experiments was recorded with video and thermal imaging cameras.

These experiments demonstrated the thermal conditions that can be generated by a "simple room and contents" fire and how these conditions can be extended along a flow path within a structure when a wind condition and an open vent are present. Two potential tactics which could be implemented from either the floor above the fire in the case of a WCD or from the floor below the fire in the case of the external water application were demonstrated to be effective in reducing the thermal hazard in the corridor.

Laboratory Experiments: Test 1. No audio.

The first fire experiment in the structure was different from the other experiments in that no external wind was being imposed to the structure. In this experiment the door between the hall and the target room was a hollow core wood door. The trash container was remotely ignited and the fire was allowed to grow. After the window was broken (vented) by the fire, a researcher in full PPE cleared the window opening with a pike pole. After the window was vented, the fire was given time to respond to the change in ventilation. After the fire within the structure was determined to be fully developed, the fire was then suppressed by safety sprinklers installed in the structure and by a manual hose stream.

At 0 seconds, ignition begins. At 60 seconds, a visible smoke layer appears. At 213 seconds, the window partially fails (pieces are missing). At 248 seconds, the window is manually vented. At 268 seconds, hot gas flows to floor in corridor IR. At 348 seconds, the target room door fails. At 493 seconds, suppression begins.

Laboratory Experiments: Test 2. No audio.

The second experiment in the series was conducted to examine the impact of wind on the structure fire and quantify the impact of the large wind control device. The large wind control device measured 2.95 m (9.66 ft) by 3.66 m (12.0 ft). In the wind control experiments, as described in Section 4.3.2, the wind control device reduced the velocity in the structure to zero. The fan speed used in this experiment was 2500 RPM, which provided a 6.7 m/s to 8.9 m/s (15 mph to 20 mph) wind speed at the window opening. A trash container fuel package was ignited remotely with and electric match to start the experiment at Time = 0 s.

Ignition begins at 0 seconds. At 50 seconds, a visible smoke layer appears. At 167 seconds, the window is mostly vented. At 169 seconds, hot gas flows to floor in corridor IR. At 180 seconds, the window is completely vented. At 201 seconds, WCD turns on. At 255 seconds, WCD is part off. At 271 seconds, WCD is off. At 293 seconds, suppression begins.

Laboratory Experiment: Test 7. No audio.

The seventh experiment in the series was conducted to examine the impact of wind on the structure fire, the impact of closing the doorway from the living room to the corridor, and to quantify the impact of a smooth bore water stream into the bedroom. The experimental preparations were made as described in Section 4. The fan speed used in this experiment was 1500 RPM, which provided a 3.0 m/s to 4.0 m/s (7 mph to 9 mph) wind speed at the window opening. A trash container fuel package was ignited remotely with and electric match to start the experiment at Time = 0 s.

At 0 seconds, ignition begins. At 200 seconds, a visible smoke layer appears. At 297 seconds, the window is partially vented. At 310 seconds, the window is cleared. At 377 seconds, the door opens. At 435 seconds, the hose at the ceiling is on. At 505 seconds, the hose is sweeping the ceiling. At 538 seconds, the hose is off. At 545 seconds, manual suppression begins. At 550 seconds, the fire is knocked down.

Laboratory Experiments: Test 8. No audio.

The eighth experiment in the series was conducted to examine the impact of wind on the structure fire, quantify the impact of a smooth bore water stream into the bedroom, and to compare results to experiment 7 (section 5.7) with the living room to corridor door open. The experimental preparations were made as described in Section 4. The fan speed used in this experiment was 1500 RPM, which provided a 3.0 m/s to 4.0 m/s (7 mph to 9 mph) wind speed at the window opening. A trash container fuel package was ignited remotely with and electric match to start the experiment at Time = 0 s.

At 0 seconds, ignition begins. At 40 seconds, a visible layer of smoke appears. At 141 seconds, the window is vented. At 145 seconds, hot gas flows to the floor in corridor IR. At 298 seconds, the hose is on at the ceiling. At 350 seconds, the hose is sweeping stream. At 435 seconds, the fire is knocked down. At 470 seconds, the hose is turned off. At 475 seconds, the hose is on at contents. At 489 seconds, the fire is out.