3.1. Test Devices
In order to accurately reproduce a real fire in an underground confined space and explore the characteristics of liquid nitrogen fire extinguishment therein, a liquid nitrogen fire extinguishing test system was built, as shown in
Figure 1. The setup consists of three basic components: a reduced-size underground pipe gallery model, a nitrogen injection system, and a data acquisition system.
The underground pipe gallery model is 10.0 m long, and its cross-section is 0.9 m wide and 1.25 m high. The whole model is made of a high-temperature resistant stainless-steel material. Specifically, the top, bottom, and the back are 3 mm thick stainless-steel plates. The front and the rotating doors on both sides of the model are made of fire-resistant glass that can withstand temperatures of hundreds of degrees Celsius, which facilitates the real-time observation of combustion, smoke spread, nitrogen injection and fire extinguishing process during the test.
The main purpose of the test was to obtain the changes in the flame height, temperature, and oxygen volume fraction in an urban underground pipe gallery under the action of liquid nitrogen. Therefore, because of the flame-retardant characteristics and the complex combustion characteristics of cables, a pool fire with an obvious combustion development stage, stable stage and decline stage was selected to simulate the fire scenario in an urban underground pipe gallery. The fire source was a square fuel-pan with a side length of 0.35 m, which was placed in the middle of the pipe gallery model. The fuel selected was n-heptane. The combustion of n-heptane is stable and only produces a small amount of smoke, which is convenient for test observation and flame height recording.
The nitrogen injection system was composed of a liquid nitrogen tank, a Coriolis Mass Flowmeter and a liquid nitrogen transmission pipeline. The liquid nitrogen tank is a self-pressurized heat-insulation tank, and the top is equipped with a liquid outlet valve and a gas outlet valve. The pressure in the tank is displayed on the pressure gauge, and the outlet pressure can be quickly adjusted with the booster valve. The mass flow rate is precisely displayed by the Coriolis Mass Flowmeter in real time. The liquid nitrogen transmission pipeline is a low-temperature and high-pressure resistant stainless-steel hose, wrapped with an insulating cover to reduce the liquid nitrogen consumption and cooling capacity loss in the transmission process. The nitrogen injection was positioned on the ceiling of the pipe gallery, and was about 0.3 m away from the fire source, which not only ensured the effect of extinguishing the fire, but also effectively prevented fuel spillage due to the direct impact of the nitrogen injection on the fire source. The nitrogen injection pipe was inserted vertically downward into the pipe gallery for about 0.1 m.
The data acquisition system included a K-type thermocouple, an oxygen volume fraction sensor, a thermal imaging camera and a high-definition camera. The K-type thermocouple was arranged on the ceiling of the pipe gallery to obtain the temperature change directly above the fire source. The oxygen volume fraction sensor was positioned near the fire source to obtain the change in the oxygen volume fraction, so as to analyze the asphyxiation effect of nitrogen injection on the fire source. In order to prevent high-temperature damage to the sensor, the oxygen volume fraction sensor was placed at the same height as the fire source and was about 0.1 m away from the fire source. The infrared thermal imager and high-definition camera were placed inside the pipe gallery and 2.5 m away from the fire source, to obtain information about the changes in the flame during the test. All data acquisition devices were connected to data lines, which were connected to computers through a hole in the underground pipe gallery model. After the data acquisition system was tested, all holes and chinks were filled with high-temperature resistant foam material to ensure the airtightness of the underground pipe gallery model.
3.2. Test Method
In order to explore the influence of the nitrogen injection flow rate and pipe diameter on the fire extinguishing efficiency of liquid nitrogen, the nitrogen injection flow rate and pipe diameter were selected as the variables to perform the nitrogen injection fire extinguishing test. In the test, the nitrogen injection flow rate was controlled by properly adjusting the degree of opening of the liquid nitrogen tank valve, and the maximum flow rate was 315 kg/h. The pressure of the liquid nitrogen tank was basically maintained at 0.09–0.1 MPa. The diameter of the nozzle used for the nitrogen injection was 10–25 mm. At the same time, the self-extinguishing test results without nitrogen injection were selected as the control group. Three repeated tests were conducted for test condition 2 and the time of the fire extinguishment was 24 s, 23 s and 29 s. Good repeatability of the test result was obtained, and the test with the time of fire extinguishment of 24 s was used for subsequent data analysis. For other tests, one test was conducted. All the test conditions are listed in
Table 1.
In the test, 0 s was the time that the test started, and 120 s was the time when the fire entered a fully developed stage. The time of 120 s was selected as the time that the nitrogen injection started (t0). When the flame went out, the time of the flame extinction (t) was recorded, the nitrogen injection was stopped immediately, and the test was over. The flame height, temperature, and oxygen volume fraction were recorded during the test.
3.3. Test Phenomenon
Taking condition 2 as an example, the flame morphology changes during the test are displayed in
Figure 2. At 0 s, the fuel pool was ignited with an extended flame spray gun, and the fire source began to burn violently. As the burning proceeded, the fire gradually expanded and the height of the flame continued to increase. At 10 s, the flame height reached the middle of the pipe gallery (about 0.5 m). At 54 s, the height of the flame exceeded 1 m, which started to affect the ceiling, and the impact frequency gradually increased. Meanwhile, half of the pipe gallery was filled with the smoke. At 120 s, the top of the flame was almost at the same height as the pipe gallery model, and the ceiling was continuously impacted by the flame. The flame and the fuel pool were basically surrounded by the hot smoke, indicating that the fire had entered the fully developed stage. At this time, the nitrogen injection system was turned on to extinguish the fire.
After the nitrogen injection valve was opened, part of the remaining air in the pipeline was driven at high speed by the nitrogen toward the pipe gallery, strongly disturbing the air flow around the fire source. At 121 s, the flame was inhibited under the impact of nitrogen, and the flame height decreased rapidly, while the flame width and range became larger. At 135 s, the proportion of liquid nitrogen in the injected two-phase nitrogen gradually became greater than that of gaseous nitrogen, and the jet column could be clearly seen at the nitrogen injection port. Under the continuous pressure of the liquid nitrogen on the flame, the flame height was always less than 0.4 m and it fluctuated from side to side below the 0.4 m horizontal line; this lasted for about 6 s. At 144 s, the flame was completely extinguished under the action of the nitrogen injection. At this time, the nitrogen injection valve was closed and the test ended.