1. Introduction
According to the 2022 Annual Statistical Report of the National Fire Agency [1], electricity has been the second most frequent cause of fire since 2012 after negligence, accounting for approximately 23% of fires. In the event of an electrical fire, the wire acts as a mediator to spread fire across a wide range, and as the amount of heat released by fire is large, the fire may spread to other combustibles. A wire is composed of a conductor with electric current, an insulator, and a jacket to prevent the leakage of electric current flowing through the conductor. The wire insulator and jacket are the main combustibles, and because of their low heat-resistant temperatures, ignition occurs readily from external fires. Additionally, as they consist of polymer composites, their combustion releases large amounts of smoke and toxic gas, posing a high risk of fire.
Smoke released upon fire is the main factor that interferes with evacuation by reducing visibility for evacuees. According to the 2021 Statistical Report on Electricity Accidents of the Korea Electrical Safety Corporation [2], electrical fires occur most commonly in housing facilities, followed by highly populated areas including industrial and residential facilities. To ensure safe evacuation in a fire event in a densely populated area, it is essential to know the smoke characteristics. Among the various characteristics, smoke density indicates the level of smoke opacity according to the measurement of the relative transmittance of light. As it is directly associated with the visibility of the evacuee, smoke density should be quantitatively measured. Various factors, such as the color and scale of smoke, affect the smoke density, and depending on these factors, light transmittance varies and visibility is determined. Hence, smoke density most reliably represents the smoke characteristics.
Numerous studies have investigated the risks of heat and smoke with regard to wire insulators and jackets. Most of these studies focused on thermal risks according to aging deterioration of wire insulators and jackets as the main cause of electrical fire. Kim et al. (2018) induced differences in the equivalent lifetime of IV insulated wires commonly used in aging buildings and reported that the risk of overcurrent increased as the equivalent lifetime increased to reduce the time of onset of carbonization and smoke [3]. Park et al. (2015) examined the cover jacket of a power cable used in confined environments of public underground structures and reported that the total amount of smoke released and toxicity index increased as the number of years of jacket use increased, i.e., as the level of aging deterioration increased [4]. In their investigation of non-Class 1E cables used in nuclear power plants, Kim et al. (2020) reported that aging deterioration increased the risk of fire, while the indicators of smoke risk, including the smoke index and the total release of smoke, tended to increase. In addition to the aging deterioration of wire jackets, many studies have focused on the deterioration of the jacket material [5]. Park et al. (2004) estimated the activation energies of insulator PVC and heat-resistant PVC to show that the latter exhibited higher values as well as long-term stability in predicting the lifetime [6]. Zhang et al. (2019) conducted a cone calorimeter test on the insulator (xylenol/spandex/nitrocellulose/Flexibilizer) of aerial cables and reported that the rate of smoke release, smoke parameters, and burned area increased in the early stage of combustion and then tended to decrease gradually. Additionally, the concentration of CO in smoke particles slowly increased in the early stage of combustion and then tended to decrease, whereas the levels of CO2 and O2 continued to increase [7]. Kim et al. (2013) investigated the deterioration of tray cables consisting of FR-PVC jackets and cross-linked polyethylene (XLPE) insulators, where the thermal dissociation temperature was measured. The net calorific value was 0.5 kW at maximum, and when the net calorific value of the initial 5 min of ignition was differentiated, the resulting total thermal energy was 24800 kJ/kg [8]. You et al. (2019) tested the combustibility of plastics (PVC, PP, PC, PS, and FRP) as the main jacket material and reported that the PVC plate had the lowest heat release among the plastic materials examined, while the PS plate exhibited the highest value. The PS plate also exhibited the highest release of CO and CO2 [9]. As such, previous studies have explored the thermal characteristics of the wire, but few studies have determined the impact of smoke released during wire combustion on evacuation.
In this study, therefore, the insulator PVC in insulated wires (formerly, IV) used widely in interior wiring of housing facilities with the highest incidence of electrical fire and the insulator XLPE of low-toxicity flame-retardant cross-linked polyolefin (HF-IX) was selected as test materials, and the changes in smoke density during combustion and the smoke density in the early stage of fire were measured, for comparing the risks in evacuation. Furthermore, the levels of light attenuation and visibility were estimated to quantify the smoke density and examine the characteristics of smoke density and combustion upon fire.
2. Experimental Methods and Conditions
2.1 Test materials
Wires used in the interior wiring of housing facilities were investigated in this study to compare the smoke density and combustion characteristics of wire insulators. Among the components of the PVC (formerly, IV) insulated wire that used to be the most common type in interior wiring and those of the HF-IX wire, which is the most popular one today, the insulator as the combustible material in each wire was selected as the test material. Table 1 presents the wire types used in this study and the insulator materials in each tested wire, along with the experimental conditions. The insulator of the PVC insulated wire is PVC, and that of the HF-IX wire is XLPE. PVC is a thermoplastic material with high moisture resistance as well as excellent weather-proof and flame-retardant properties. Under the influence of thermal energy, however, deterioration is accelerated by dehydrochlorination. XLPE has superior thermosetting and viscoelasticity properties and thus exhibits outstanding mechanical and electrical properties as well as excellent flame retardancy, water resistance, and abrasion resistance. For the reliability of experimental data, the mass was set constant at 15 ± 1 g, and the samples were tested after 24 h of pretreatment at 23 °C and 50% relative humidity. The 75 mm × 75 mm square samples were wrapped with aluminum foil to prevent residual attachment to the scaffold after the experiment, and as a result, the samples were prepared so that the area of exposure to radiation heat inside the scaffold was 65 mm × 65 mm. The samples were arranged parallel to the lower end of the cone heater, and according to the test conditions, 25 kW/m2 of heat was applied to the surface of the sample depending on the presence or absence of the ignition flame. According to the standard criteria ISO 5659-2 [10], the expandable PVC sample was positioned at 50 mm on the lower end of the cone heater, and the non-expandable XLPE sample was positioned at 25 mm, where the radiation heat was applied.
2.2 Experimental device
The purpose of this experiment was to quantitatively determine the risk of smoke released by each test sample upon fire. The risk of smoke in relation to evacuation is quantified as visibility, and an experimental device based on optical equipment to estimate smoke density is necessary to determine visibility. Thus, a smoke density chamber was used to measure optical density according to the standard methods of ISO 5659-2, and a schematic of the chamber is shown in Figure 1. The size of the smoke density chamber is 0.914 m × 0.914 m × 0.610 m (width × length × height), with a total volume of 0.51 m3, and the chamber consists of a cone heater, an ignitor, a load cell, an optical window, a light source, and optical equipment. The cone heater radiates 25 kW/m2 of heat to the test sample surface from distances of 25 and 50 mm. The ignitor releases the ignition flame by mixing propane gas and oxygen. The ignition flame is released 15 mm below the cone heater, while it serves as the source of ignition during the combustion. The load cell is a device that measures the mass loss; it measures the changes in mass during the experiment, whereby the mass optical density as an indicator of smoke density can be obtained. The accuracy of the load cell is 0.5 g, and in this study, the mass data from the load cell were not analyzed, as the device was determined not to be adequately sensitive to the reduction in mass. There is an optical window on each of the upper and lower sides of the chamber, and from a 9-W light source installed below the optical window on the lower side, a ray of light is sent toward the optical window on the upper side. The light intensity along the light path is measured by the optical equipment on the upper side. The light from the source passes the neutral density (ND) filter inside the optical equipment, and the ND filter equalizes the smoke wavelength and color intensity to uniformly attenuate the transmittance of specific rays of light to reduce errors in optical density measurements.
2.3 Specific optical density
Each test sample was prepared as a 75 mm × 75 mm square and wrapped with aluminum foil to prevent residual attachment to the scaffold after the experiment. When the sample was placed in the scaffold, the area exposed to radiation heat was 65 mm × 65 mm. The samples were arranged in parallel to the lower end of the cone heater, and according to the test conditions, 25 kW/m2 of radiation heat was applied to the sample depending on the presence or absence of the ignition flame. On the basis of the standard criteria, the expandable PVC sample was positioned at 50 mm on the lower end of the cone heater, and the non-expandable XLPE sample was positioned at 25 mm, where the radiation heat was applied. The vertically set optical equipment measures the intensity of light passing through it, and the measured intensity is converted into transmittance to obtain the smoke density. After the conversion, the transmittance data can be used to estimate the optical density as an indicator of the smoke density. In this study, the specific optical density Ds was measured, as it reflects the morphological characteristics of the smoke density chamber. Ds indicates the level of opacity of the smoke, and it is expressed by Eq. (1). To obtain Ds, the logarithm of the relative transmittance of light was calculated, and the total volume of the smoke density chamber was multiplied by the volume against the vertical length of the chamber occupied by the test sample, which represents the morphological characteristics of the chamber. The mean of triplicate measurements was calculated for each test condition.
Ds is obtained by multiplying G, which describes the shape of the smoke density chamber, by the optical density. G is expressed as V/AL, where V represents the volume [m3] of the chamber, A represents the exposed area [m2] of the test sample, and L represents the length [m] of the transmitted ray of light. For calculating the optical density, T represents the transmittance of the ray of light, and F represents the density of the optical filter; the value is 0.3 if an optical filter is used and 0 if an optical filter is not used.
3. Results and Discussion
3.1 Changes in optical density over time
Figure 2(a) shows the specific optical density Ds of the PVC insulated wire. The maximum Ds (Ds max) of the PVC insulated wire is 242.3 in the presence of the ignition flame and 476.1 in the absence of the ignition flame. These large values are due to the combustion characteristics of PVC—the key material of the PVC insulated wire. The molecular formula of PVC is (C2H3Cl)n, with a Cl atom; thus, Cl radicals are generated upon combustion [11]. As the Cl radicals react with O and H radicals, a large amount of smoke is released through incomplete combustion of PVC [12], generally leading to a high Ds of the PVC insulated wire. In the flame mode with the ignition flame, the flame is ignited on the surface of the sample within a minute and is extinguished in approximately 7 min. Upon ignition, Ds increases rapidly owing to the sensitivity of the reaction rate to temperature with the high activation energy of PVC, i.e., 282.05 ± 12.22 kJ/mol [12]. The rapid increase in sample temperature accompanying ignition leads to a high-rate combustion reaction, accelerating the process of degradation, which causes the rate of smoke release to increase rapidly [13]. Additionally, dehydrochlorination occurs at approximately 250 °C during the combustion of PVC, leaving polyene [14]. At temperatures above 400 °C, polyene decomposes into gas, oil, and char, with the yield of char increasing as the temperature increases [12]. Char is characterized by its ability to protect the interior of the test sample from external radiation heat [15]. In the presence of the ignition flame, the increased rate of degradation with ignition causes a rapid increase in Ds, and owing to the resulting rapid degradation of the sample, the amount of combustible decreases with the formation of char so that the smoke density gradually decreases. Hence, the time difference in reaching the point at which a fall in Ds is detected is attributable to the low yield of char and low rate of degradation due to the low temperature in the non-flame mode, compared with the flame mode.
Figure 2(b) shows the Ds of the HF-IX insulator. As with the PVC insulated wire, the Ds in the non-flame mode is higher than that in the flame mode for the HF-IX insulator. However, compared with the PVC insulated wire, the Ds in the non-flame mode for the HF-IX insulator is considerably lower. This is because the halogen-free flame retardant (HFFR) added for the low-toxicity flame-retardant properties of the HF-IX insulator produces water and carbon dioxide only upon combustion, so that negligible amounts of carbonized materials—the main cause of reduced visibility—are formed [16]. In the flame mode, the flame is ignited on the surface of the sample within 4 min and is extinguished in approximately 7 min. In contrast to the PVC insulated wire, the Ds does not increase rapidly upon ignition. The lack of a sudden change in Ds despite the rising temperature is due to the water released from the aluminum hydroxide used as the HFFR upon direct heat exposure, which protects the test sample from the heat [17]. The surface of the HF-IX insulator after the experiment is presented in Figure 3, for the flame and non-flame modes. The surface of the test sample is shown to be damaged to a higher extent in the flame mode with direct flame exposure caused by ignition, leading to the release of water from the aluminum hydroxide.
Table 2 presents the Ds max measured in each condition and the time taken to reach Ds max for the test samples. Regardless of the presence or absence of the ignition flame, the PVC insulated wire exhibited a higher Ds max than the HF-IX insulator. This implies that the risk of smoke is higher for the PVC insulated wire than the HF-IX insulator. Comparing the quantified risk of smoke reveals that the Ds max measured in the flame mode was approximately 10 times higher for the PVC insulated wire with a higher risk of smoke than for the HF-IX insulator with a lower risk of smoke, while the time taken to reach Ds max was approximately 8.3 times higher. Additionally, the Ds max measured in the non-flame mode was approximately 2 times higher for the PVC insulated wire, with an approximately 1.5-fold shorter time to reach Ds max, compared with the HF-IX insulator.
3.2 Extinction coefficient and visibility
Figure 4 shows the changes in the extinction coefficient Cs with respect to time for each wire, and Table 3 presents the Cs max measured for each wire. As a quantified value of light transmittance with respect to the concentration of smoke, Cs serves as an indicator of the level of light attenuation. Cs decreases as the amount of transmitted light decreases. According to Bouguer's law of monochromatic light attenuation, Cs is expressed as follows:
Cs is obtained by multiplying the natural logarithmic values of the ratio of transmittance and L, i.e., the distance between the light source and the photoreceptor. T0 represents the transmittance in the absence of smoke, and T represents the transmittance in the presence of smoke.
Using Cs, the visibility ensured by the Ds measured upon sample combustion can be determined; thus, Cs is used as an indicator that allows a visual comparison of the risk of smoke. Table 4 presents the visibility ensured for different values of Cs [18]. Cs max is 6.2 m-1 in the absence of the ignition flame and 4.5 m-1 in the presence of the ignition flame for the PVC insulated wire, and it is 4.5 m-1 in the absence of the ignition flame and 0.4 m-1 in the presence of the ignition flame for the HF-IX insulator. The measured Cs max is 4.5 m-1 for the PVC insulated wire in the flame and non-flame modes and the HF-IX insulator in the non-flame mode, with the exception of the Cs max measured for the HF-IX insulator in the flame mode. Hence, the visibility ranges from 0.2 to 2 m. This is far below the 5-m visibility criteria for the evacuation threshold of smoke concentration regarding the residents of buildings to represent a characteristic of housing facilities.
3.3 Changes in initial smoke density VOF4
Evacuation should be performed in the early stage of fire; thus, it is difficult to ensure adequate evacuation time when the initial Ds upon fire is high. It is thus necessary to compare the initial smoke densities for different test samples. In the early stage of combustion, the trend of the smoke density is more effectively determined by comparing cumulative values than by comparing point values; hence, the VOF4 as an approximation of cumulative optical density measured for 4 min in the early stage of combustion was estimated. VOF4 is the cumulative value of Ds obtained under the assumption that it tends to increase by y = x based on the mensuration by parts, and it is expressed by Eq. (3).
Ds(1), Ds(2), Ds(3), and Ds(4) are the Ds values after 1, 2, 3, and 4 min of exposure of the material to radiation heat.
Table 5 presents the values of Ds, VOF4, and Cs for the respective VOF4 by time. For the PVC insulated wire, the Ds max was higher in the non-flame mode than in the flame mode, whereas the VOF4 was approximately 4.1 times higher in the flame mode. This implies that evacuation in the case of the PVC insulated wire is easier in the absence of the ignition flame in the early stage of fire, although the high Ds max indicates a high risk of smoke. The standard guideline of performance-based design, test, and operation of such facilities as the fire protection system states that the permitted time of evacuation is 4 min in the case of fire in a mid-/high-rise building where a recorded voice message or warning announcement by a trained firefighter is available. The Cs for the VOF4 measured for the PVC insulated wire in the flame mode is 3.9 m-1, indicating a visibility range of 0.2 to 2 m. As the population density is high in housing facilities with a small number of assigned firefighters, evacuating from a mid/high-rise residential building within 4 min in the visibility range of 0.2-2 m upon fire is unlikely to be easy.
For the HF-IX insulator, both Ds max and VOF4 are low in the flame mode. This implies that evacuation from the smoke arising in the presence of the ignition flame is easier in the case of PVC insulators in the early stage of fire, while the low Ds max indicates a low risk of smoke. While VOF4 is higher in the non-flame mode than in the flame mode, the visibility range at the respective VOF4 is 20-30 m, which does not pose significant problems with the same evacuation condition used in the flame mode. Hence, the evacuation in the early stage of the fire is likely to be easy with a stable visibility level of smoke for the HF-IX insulator.
4. Conclusions
The effects of the presence of the ignition flame on the smoke density for the insulator component of the wire were analyzed in accordance with the ISO 5659-2 combustion chamber method. The following conclusions were drawn.
1. The specific optical density Ds is closely associated with the characteristics of combustion and thermal degradation of the test sample; it rapidly increases upon surface ignition in the case of PVC with a high activation energy. In the case of XLPE, where the aluminum hydroxide added for the low-toxicity flame-retardant function releases water upon direct exposure to heat to protect the test sample, no sudden increase in Ds is observed even upon ignition.
2. For PVC, evacuation in the early stage of fire is difficult in the presence of the ignition flame, while the risk of smoke is low. In contrast, for XLPE, evacuation in the early stage of fire is easy in the presence of the ignition flame, with an equally low risk of smoke. This suggests that the risk regarding the maximum and initial Ds may vary according to the given conditions upon evacuation.
In this study, a combustion experiment was performed on wire insulators, and the resulting data were used to estimate Ds and identify the cause of the changes in Ds in line with the chemical changes in the materials during combustion. By comparing the maximum and initial Ds values for each insulator through VOF4, the risks after a set period of time post-fire were assessed for each insulator, and the differences upon evacuation in the early stage of fire were discussed.
PDF Links
PubReader
ePub Link
Full text via DOI
Download Citation
