Experimental Study on Particle Size Distribution Characteristics of Aerosol for Fire Detection

Abstract

Current optical fire smoke detectors use scattering light intensity as an indicator of smoke concentration and trigger fire alarms when the intensity exceeds a threshold value. However, such detectors are prone to false alarms caused by non-fire aerosols since both fire smokes and non-fire aerosols scatter light. Thus, in order to reduce false alarms caused by non-fire aerosols such as dust and water vapor, fire detectors must be capable of distinguishing fire smoke from non-fire aerosols. Since the light scattering signals depend on the particle size information of aerosols, it is essential to study and characterize the particle size distribution of fire smoke and non-fire aerosols for differentiating them. In this paper, a comprehensive aerosol experimental platform is built to measure the particle size distribution of various typical fire smokes and non-fire aerosols. Through the conducted experiments, we note that there are significant differences in the particle size distributions of typical fire smokes and non-fire aerosols, with a boundary of about 1μm. The experimental results provide fundamental data support of the particle size distribution for developing a better fire detector that accurately identifies smoke.

1. Introduction

Photoelectric fire smoke detectors measure the intensity of light scattered by particles to determine whether a fire has occurred in the monitoring area. This type of detector is widely used since the operating principle is simple to implement and has a high sensitivity. However, non-smoke particles (such as dust, water vapor, oil mist, etc.) can also scatter the light and trigger false fire alarms [1,2,3,4,5].

Many works have been conducted to improve the alarm accuracy of photoelectric fire smoke detectors using multi-wavelength and multi-channel light scattering signals [6,7,8,9,10]. Philipp et al. designed a dual-wavelength, three-angle aircraft-specific fire detector based on infrared and green light. The detector distinguishes between fire and non-fire aerosols by the ratio of the three light scattering signals to each other according to the experimental database in the aircraft cabin [11,12]. Some detectors not only acquire aerosol concentration information, but also extract the statistical characteristics of aerosols from multi-channel light scattering signals to assist in identifying smoke, such as the volume concentration and surface area concentration of particles [13,14,15]. The essence of these studies is to construct a nonlinear optical fingerprint of smoke and non-fire aerosols based on their optical parameters. To obtain the optical characteristics of smoke aerosols, Loepfe et al. combined multi-angle and multi-wavelength to test the light scattering characteristics and parameters such as the particle size of fire smokes in the European standard fire test (EN54) [16,17]. Hyo-Yeon Jang et al. analyzed the differences of scattering light with different angles and wavelengths for identifying the type of smokes [18].

According to the Mie theory, the particle size distribution of aerosols is one of the most important influencing factors of their optical characteristics [19,20,21,22,23]. Therefore, it is essential to study the particle size distributions and the differences of typical fire smokes and non-fire aerosols for developing fire detectors with higher alarm accuracy. Krull et al. measured the particle size distribution and rise rates of typical interference aerosols during sanding wood, grinding red brick, blowing cement, and so on [24,25]. On the other hand, the particle size distributions of standard fire smokes are characterized by Ma et al. using SMPS (Scanning Mobility Particle Sizer, TSI SMPS-3936, TSI Incorporated, St. Paul, MN, USA) [26,27,28]. A further study to extract the difference between fire smokes and interference aerosols needs to be conducted.

In this paper, an aerosol particle-sizing platform is set up to study the particle size distribution of both typical fire smoke and non-fire aerosols. Thus, the particle size distributions of various typical fire smoke and non-fire aerosols are measured. By combining the experimental data and the related literature, this paper analyzes the differences in the particle size distribution between fire smoke and non-fire aerosols, which provides the basic knowledge for fire smoke identification.

2. Materials and Methods

In order to exclude interference from non-fire aerosols, an aerosol experimental platform is set up to study differences in the particle size distributions of typical fire smokes and non-fire aerosols, as shown in Figure 1. According to the block diagram illustrated in Figure 2, the platform consists of an aerosol generator and storage module and a particle sizing module.

In the aerosol generator and storage module, there are three different types of aerosol chambers for fire smokes, dust samples, and spray aerosols, correspondingly. Fire smokes are generated by corresponding smoke generators and stored in a 50 cm × 50 cm × 50 cm smoke chamber. Specifically, smoldering cotton and smoldering wood are heated by an electric heater, where the power of the heater is well controlled to ensure that the combustibles are always in a state of smoldering combustion; open fire of Polyurethane and open fire of N-heptane is held in a heat-resistant, wide-mouth container and ignited with open fire. A pair of 12 cm diameter fans are installed diagonally on the smoke chamber to ensure that the smoke concentration is evenly distributed throughout the chamber. An air bag, which is connected to the external air through a pipe, is equipped to ensure the air pressure balance in the smoke chamber and the stable concentration of smoke particles during aerosol sampling procedure.

Dust experiments are conducted in a transparent sphere of 0.3 m radius. The standard dust is held on the bottom of the sphere chamber, where a turbine fan with 13,500 rpm is installed at a 10 cm height from the bottom to blow up the dust. The sampling port at the top of the spherical chamber is connected to the particle size measurement equipment for dust sampling.

The fog cabinet is used to simulate water mist interference. This equipment uses high-pressure airflow to atomize the water and release it from the nozzle to the whole box. An exhaust port at the back of the cabinet is used to sample the water mist to the aerosol particle size analysis system.

In the aerosol particle sizing module, SMPS, APS (Aerodynamic Particle Sizer, TSI APS-3321, TSI Incorporated, St. Paul, MN, USA), and HELOS (Helium-Neon Laser Optical System, Sympatec HELOS/RODOS, Sympatec GmbH, Clausthal-Zellerfeld Germany) are used to obtain the particle size distribution information for typical fire smokes and non-fire aerosols. The measurement range of SMPS 3936 is 0.02–1 μm, while that of APS 3321 is 0.5–20 μm. Although their measurement principles are different, the results of these two instruments can be merged to achieve a wide particle size range of 0.02–20 μm for measuring fire smokes and non-fire aerosols. HELOS provides non-contact measurement of the particle size distribution in real-time based on Fraunhofer diffraction, which is suitable for measuring water mist and oil spray. Depending on the lens selected, our HELOS has a measurement range of 0.1–35 μm. Thus, based on these aerosol particle size measurement instruments, our experimental platform is able to measure the particle size distribution and concentration of typical fire smoke and non-fire aerosols.

3. Results

3.1. Typical Combustion Smoke Particle Size Measurement

Since combustion usually starts with pyrolytic smoldering, it is meaningful to detect fire at this early stage. Hence, we first study the particle size distribution of smoldering smokes of typical combustibles in daily life, such as cotton wicks, beech wood, mosquito coils, paper, bamboo, rubber, plastics, and so on. The particle size distributions of natural and artificial combustible materials are shown in Figure 3a,b, respectively, from which it can be seen that the particle size distribution of smoldering smokes varies with the combustion material. Depending on the combustible materials, the peak particle size is concentrated between 100 and 300 nm. Meanwhile, the smoldering fire smokes can be well described by the lognormal model and the particle size range is generally on the submicron scale.

In addition to the smoldering smokes, we also study the particle size distributions of open fire smokes, including polyethylene, PPR (Polypropylene-Random), n-heptane, kerosene, and so on. As per the results shown in Figure 4, the particle size distribution of open fire of polyurethane and n-heptane is generally larger than that of smoke from smoldering fires, with peaks concentrated between 200 and 400 nm and most particles smaller than 1 μm. Since the maximum range of SMPS is 1 μm, the particle size distribution was truncated at 1 μm. The APS measurements show that the proportion of particles larger than 1 μm is very low.

The particle size distribution of smoke particles is fitted according to the experimental data in Figure 3 and Figure 4. It can be observed that the particle size of smoke particles typically follows the lognormal distribution.

Table 1. Statistical parameters of particle size distribution of typical fire smokes.

Combustibles	Median Particle Size (nm)	Standard Deviation
Cotton wicks	123.0	1.74
Fir	177.3	1.65
Bamboo	104.0	1.84
Cotton fabric	173.0	1.69
Cardboard	229.9	1.67
Polyester	128.8	2.08
Papers	272.2	1.69
Woolen thread	80.6	1.52
Nylon	96.3	1.55
PVC (Polyvinyl chloride)	263.2	1.77
Insulation rubber	137.6	1.99
Polystyrene	80.2	1.93
Mosquito coil	165.1	1.85
Polyethylene	180.3	1.54
Kerosene	300.8	1.70
PPR	278.3	1.55
N-Heptane	308.5	1.45
NBR/PVC (Nitrile Butadiene Rubber/Polyvinyl chloride)	176.8	1.53

summarizes the statistical characteristics parameters of the particle size distribution of fire smoke.

3.2. Typical Non-Fire Disturbance Aerosol Particle Size Distribution Measurements

As the most common non-fire interference aerosols, standard dust (JIS Z 8901 Class 11, ISO 12103-1 A1, and ISO 12103-1 A2), water mist, and oil spray are studied by our platform. The particle size distributions of standard dust samples are on the micron scale, mostly ranging from about 1 to 10 μm. According to the datasheets of standard dusts, the order of particle sizes from small to large is JIS Z 8901 < ISO 12103-1 A1 < ISO 12103-1 A2. The results in Figure 5 show a similar trend among different types of dust samples, but the particle size distributions are all smaller than the reference described in the datasheets because larger particles are more difficult to blow up and settle faster. It is similar to the real operation condition where fire smoke detectors are usually fitted to the ceiling and only blowing dust particles can enter the detector chamber.

Water mist is another common source of false alarms, especially prone to trigger false alarms in hotels and kitchens. We tested the particle size distribution of different types of spray on our platform, including high-pressure gas sprays (fire test spray, sunscreen sprays, gels, freshener, and Yunnan Baiyao spray), manually pressed sprays (perfumes, mosquito repellents), and water sprays generated by the fog cabinet and humidifier. The results are shown in Figure 6. It can be seen that most of the spray particles are in the micron range, similar to dust.

The particle size distribution of typical interference aerosols was fitted according to the experimental data in Figure 5 and Figure 6. It can be seen that the particle sizes of these interfering aerosols generally follow the lognormal distribution.

Table 2. Statistical parameters of particle size distribution of typical interference aerosols.

Name	Median Particle Size (µm)	Standard Deviation
A2	3.2198	1.5129
A1	2.4152	1.5394
JIS	1.4896	1.4801
Fire test spray	7.2575	1.9974
Sunscreen spray	41.726	1.3151
Gel	3.3891	1.1598
Yunnan Baiyao	27.358	1.4405
Perfume	43.772	1.2347
Freshener	39.816	1.5268
Humidifier	9.4085	1.9872
Mosquito repellent perfume	39.902	1.4046
Water spray (fog cabinet)	4.2877	1.2939

summarizes the statistical characteristic parameters of the particle size distribution of typical interference aerosols.

4. Discussion

The comparison from Figure 3 to Figure 6 shows that there is a huge difference in the distribution range of fire smoke and non-fire aerosols. Most of the fire smoke particles are concentrated at the submicron level, with the particle sizes ranging from tens to hundreds of nanometers, and most smoke particles have no particle or only a small number of particles above 1 μm. In contrast, the particle size of most non-fire aerosol samples is in the micron scale and concentrated in the range of 1–10 μm. Comparing the particle size distribution of fire smokes and non-fire aerosols, it is obvious that these two kinds of aerosols have different particle size distribution ranges. Specifically, the particle size of non-fire aerosols is generally much larger than that of fire smokes. According to the Mie scattering theory, non-fire aerosols with larger particle size tend to have a stronger light scattering ability than small fire smokes, which would trigger false alarms of photoelectric fire detectors more easily.

The experimental results indicate that typical fire smokes and interference aerosols can be differentiated around 1 μm. Although some of the dust particles contain particles smaller than 1 μm, the proportion of sub-micron particles in these aerosols is very low. As discussed above, fire smoke and non-fire aerosols have different particle size distribution characteristics, which are closely related to light scattering characteristics.

5. Conclusions

The particle size distribution is closely related to the light scattering characteristics of fire smokes and interference aerosols. In order to develop fire detectors with higher alarm accuracy, an experimental study on the particle size distribution is conducted for characterizing typical fire smokes and interference aerosols. A comprehensive aerosol experimental platform is set up to measure the aerosol particle size distributions. The particle size distributions of typical aerosol samples are measured in high resolution. A primarily analysis of our experimental results indicates that the particle size distribution of the fire smoke we measured is mostly on the sub-micron scale, while that of several common non-fire aerosols is mostly on the micron scale, which may be useful for discriminating fire smokes from interference aerosols in general scenarios. With detailed information on the particle size distribution, more distinguishing statistical characteristics for fire smoke identification could be studied in the future. Furthermore, the light scattering characteristics can be further inferred based on the Mie theory for improving the fire smoke identification ability of the optical fire smoke detector.