- Email: [email protected]
Development of a model to calculate asbestos fiber from damaged asbestos slates depending on the degree of damage
Journal of Cleaner Production 86 (2015) 88e97
Contents lists available at ScienceDirect
Journal of Cleaner Production journal homepage: www.elsevier.com/locate/jclepro
Development of a model to calculate asbestos fiber from damaged asbestos slates depending on the degree of damage Young-Chan Kim, Won-Hwa Hong*, Yuan-Long Zhang School of Architecture, Civil, Environmental and Energy Engineering, Kyungpook National University, 80 Daehak-ro, Buk-gu, Daegu 702-701, Republic of Korea
a r t i c l e i n f o
a b s t r a c t
Article history: Received 9 July 2014 Received in revised form 28 August 2014 Accepted 28 August 2014 Available online 6 September 2014
It is very important to quantify asbestos fiber emitted due to disasters and formulate measures to prevent scattering as to maintain, dismantle and remove asbestos slates effectively in normal times. The researchers developed a model to estimate the amount of asbestos fiber generated depending on the damage of slates, focusing on natural damages. The researchers collected used slates from typical buildings and classified them into 3 grades depending on the severity of damage. The researchers freedropped slates from the roof height of a typical building, and sampled the air surrounding it. With the air sampled for each of the 3 grades of damage, the researchers counted the asbestos fibers by using the Scanning Electron Microscopy/Phase Contrast Microscopy. By conducting a regression analysis with SPSS 20.0 based on data acquired, the researchers deduced a formula. The linear regression analysis showed that ‘damaged slate generates asbestos fiber per unit area ¼ 0.077 þ 0.159 area of slate (coefficient of determination: 70.1%)’. Generation of asbestos fiber per unit area was 0.127 f/cc (Good), 0.157 f/cc (Normal), and 0.221 f/cc (Bad). The estimation model offered in this study can quantify the amount of asbestos fiber based on the building area by estimating the amount of asbestos fiber scattered due to the slate's severity of damage. This estimation model is expected to contribute to the quantification of asbestos generated due to disasters, including earthquakes, by building and by area, which were not been possible before due to the lack of relevant studies. © 2014 Elsevier Ltd. All rights reserved.
Keywords: Asbestos Damaged asbestos cement slate Asbestos fiber Degree of damage
1. Introduction Asbestos has been used since ancient times due to its high affordability and desirable physical properties such as resistance to heat and fire as well as its antiseptic and insulating properties. With the utilization of steam engines since the Industrial Revolution, asbestos consumption increased rapidly (Becklake, 1976; Sim, M.R., 2013). Global consumption of asbestos started to increase rapidly in the 1940s and reached its peak in the 1980s (Virta, 2006). Exposure to asbestos, however, is known to cause incurable diseases after a latent period of 20e50 years such as pulmonary asbestosis, lung cancer with unfavorable prognosis and malignant mesothelioma (Doll et al., 1985; Hourihane, 1964; Linton A. et al., 2012). Harmfulness of asbestos has long been a subject of research (Wagner, 1965; Kamp, D.W., 2009). Due to this asbestos use has been reduced gradually since the 1970s and some countries have been establishing related laws to prohibit or limit development of
* Corresponding author. Tel.: þ82 53 950 5597; fax: þ82 53 950 6590. E-mail addresses: [email protected], [email protected] (W.-H. Hong). http://dx.doi.org/10.1016/j.jclepro.2014.08.092 0959-6526/© 2014 Elsevier Ltd. All rights reserved.
asbestos since the early 1990s (Kane et al., 1996; Nicholson, 2001; Selikoff et al., 1964; Deng et al., 2009). Nevertheless, asbestos is still used in large quantities in many countries as building material especially in the form of asbestos cement (Kazan-Allen, 2005; Jinhui et al., 2014). Asbestos-caused diseases generally occur through inhalation of asbestos fiber (Kane et al., 1996; Ernst, W., 2012). Asbestos cement is not deemed harmful to health since the asbestos fibers are strongly bound by the cement (U.S. EPA, 2003). However, it caused problems when asbestos fibers are lost and emitted to their surroundings through years of weathering (Bornemann and Hildebrandt, 1986). Use of asbestos is prohibited or limited in most countries but is increasing in Latin America, Russia and Asia (Kazan-Allen, 2005). Unlike the increasing use of asbestos in Asia, Korea started to import asbestos in the 1960s, it recorded its peak use in 1992 by about 95 thousand tons but has shown a consistent decrease since then (Kim et al., 2009). After inserting asbestos in the list of harmful substances that require permission to be used with the revision of the Occupational Safety and Health Act in 1990, Korea banned the use of asbestos completely with the revision of an enforcement decree of the act in 2009 (Korea ME, 2009). In addition, Korea has
Y.-C. Kim et al. / Journal of Cleaner Production 86 (2015) 88e97
89
Fig. 1. Process flow of this research.
established the Asbestos Injury Relief Act and the Asbestos Safety Management Act to regulate compensation for damage caused by asbestos and to control the safety of used asbestos (Kim et al., 2014). In Korea, about 96% of the imported asbestos in the 1970s was used in slates, a type of asbestos cement, and it was used about 82% of the time in the 1990s (J.K. Choi et al., 1998). In Korea, slates were manufactured in a mixture of about 90% cement and 10% chrysotile. All the slates produced were asbestos cement slates (Kim et al., 2010). Calcium hydroxide, a component of cement, is watersoluble (Beddoe and Dorner, 2005). As time passes, asbestos fibers contained in the slate are emitted into the surroundings (Bornemann and Hildebrandt, 1986) and solubility grows in acid rain (Dyczek, 2006). The high concentration of asbestos measured in the air around buildings with deteriorated slate roofs has a harmful influence on health (Spurny, 1989). About 80% of asbestos emitted from corroded surfaces is reported to be washed away by rain water with about 20% being emitted into the air, but the exact ratio has not been verified (Meyer, 1986). Recent research reported that moss can protect the surface of the slate as the metabolite of moss converts the chemical component of asbestos fibers into a harmless substance (Favero-Longo et al., 2009; Turci et al., 2007) however this is not an ultimate solution. Considering the global trend of banning asbestos, the situation in Korea, and the hazards of slates, the solution against asbestos-related problems in Korea should be effective maintenance, safe removal and safe disposal (Kim et al., 2011a,b). Having reviewed the studies conducted in Korea and overseas, the researchers could find a number of research on the health hazard of asbestos and emission of asbestos from slates, but could not find any research on the amount of asbestos fiber generated from damaged slates. The damage of a given slate is divided into artificial and natural damage (e.g., earthquake). This research focuses on natural damage as it is designed for application in disaster situations including earthquakes. Before starting the research, slates used in typical buildings were collected and classified into 3 grades depending on the severity of damage. To simulate natural damage, the researchers free-dropped slates from the roof height of a typical building and sampled the air surrounding it. The
researchers then counted the asbestos fibers in the sampled air depending on the severity of damage, and deduced a regression equation for the amount of asbestos fiber by the area of the slate. This study estimates the amount of asbestos fiber generated due to damage of slates by building and by district, and is expected to be widely used in policy decisions such as measures to prevent asbestos damages and development of guidelines. To deduce a regression equation for the amount of asbestos generated by the severity of damage of slates, the researchers collected used slates from typical buildings in Korea. To establish the severity of damage, the researchers compared and analyzed various methods to assess potential scattering and exposure of asbestos-containing materials such as: the US Asbestos Hazard Emergency Response Act (AHERA, 1987), the American Society for Testing and Materials (ASTM, 2014), and the UK Health & Safety Executive (HSE, 2002). As a result of the comparative analysis, the researchers established the appropriate criteria for assessment of the damage of slates, established 3 grades (A, B and C) for damage of the collected slates, selected 3 slates from each grade, and measured the area and weight of each slate. To conduct the experiment, the researchers constructed a double-walled vinylsealed chamber and free-dropped the slates of each grade from the roof height of a typical building. The asbestos fibers generated from the damaged slate in the chamber were sampled with the asbestos samplers and mixed cellulose ester membrane filter (MCE), counted with the Scanning Electron Microscopy (SEM)/Phase Contrast Microscopy (PCM), and the results entered into a database. By conducting a regression analysis with SPSS 20.0 based on the database, the researchers deduced the formula to calculate the amount of asbestos fiber depending on the severity of damage of the slate. Fig. 1 illustrates the process flow of this research. 2. Method and application of theories In order to raise understanding of this study, this section describes types/characteristics and distribution of asbestoscontaining materials used in Korea, and provides the overview of a slate. To establish severity of damage, the researchers compared
90
Y.-C. Kim et al. / Journal of Cleaner Production 86 (2015) 88e97
Table 1 Specifications of slates specified in the Korean Standard. Type Small Corrugation
Large Corrugation
No. No. No. No. No. No.
6 7 8 6 7 8
Width (mm)
Length (mm)
Thickness (mm)
Average weight (kg/sheet)
Number of corrugation
Depth of corrugation (mm)
Asbestos content (%)
720 720 720 960 960 960
1800 2100 2400 1800 2100 2400
6.3
14 16 18 18 21 24
11.5
15
10
7.5
35
and analyzed various methods to assess potential scattering and exposure of asbestos-containing materials, such as the US Asbestos Hazard Emergency Response Act (AHERA), the American Society for Testing and Materials (ASTM), and the UK Health & Safety Executive (HSE), and established the criteria to assess damage of slates. Lastly, this section describes sampling and analysis methods used in this study.
2.1. Overview of a slate The asbestos-containing materials for typical buildings in Korea include slates (outdoor roof material), tex (outdoor ceiling finishing material), bamlite (indoor wall finishing material), gasket (facility joint), asbestos cloth (fire-fighting sheath), etc. (Bae et al., 2013). Slates have been used as a roof material for 1,260,000 buildings (18.09%) in Korea. When compared with indoor asbestoscontaining materials, slates are more easily dropped and damaged by external factors (vibration, wind, etc.). Damaged slates can generate asbestos fiber, which has a harmful influence to the environment and health. Therefore, research on scattering asbestos from slates has been conducted continuously. Firstly, Bornemann and Hildebrandt (1986) reported that an older slate roof which has long been installed or a slate roof with a damaged surface emits asbestos fibers to the air at an annual average rate of 3 g/m2 (Bornemann and Hildebrandt, 1986). Secondly, Spurny (1989) reported that asbestos cement slates would have corroded surfaces due to the weather changes at the rate of about 0.01e0.024 mm a year (Spurny, 1989). Thirdly, Pastuszka (2009) reported that asbestos cement slates would generate asbestos fibers at the rate of 2.7e6.9 103 F/(m2 J) when impacted (Pastuszka, 2009). So far,
however, there has been no research on the amount of asbestos fiber generated from damaged slates. Therefore, this research is aimed at the deduction of a regression equation to estimate the amount of asbestos fiber generated from a damaged slate out of asbestos-containing materials. In Korea, slates were manufactured by two companies, BYUCKSAN and KCC, in accordance with Korean standards. Therefore, all slate products used in Korea contained asbestos (Kim et al., 2011b). Slates are classified into small-corrugated slates and largecorrugated slates depending on the size of corrugation. Largecorrugated slates are normally used in plants and stalls, while small-corrugated slates are normally used in general buildings. Therefore, this research was conducted with small-corrugated slates produced in Korea. Table 1 shows specifications of the slates as per Korean standards (Kim et al., 2011a,b).
2.2. Slate damage assessment method Methods of assessing damage of asbestos-containing materials include: AHERA and ASTM of USA, and HSE of UK. The AHERA method divides the damage of asbestos-containing materials into serious damage, medium damage and good depending on the severity of damage. This method provides a simple assessment system and enables a systematic integration. The ASTM method divides the state of the samples into good, normal and bad depending on the severity of damage. This method provides scoring of each risk factor, and enables building managers and owners to actively take measures based on their own judgment. The HSE method divides the state of asbestos-containing materials into good, slight damage, medium damage and serious damage. This
Fig. 2. Criteria for severity of damage of slates.
Y.-C. Kim et al. / Journal of Cleaner Production 86 (2015) 88e97
Fig. 3. Walton-Beckett graticule.
method sets the priority order of management, assessing dustability and hazard, respectively. For this research, the researchers compared and analyzed the above three methods and deduced the appropriate grades for the damage of slates. The states of the slates are divided into grades A (good), B (normal) and C (bad) depending on the severity of damage: grade A is given for slates with no or minimum damage, B for slates with medium severity damage covering not more than 10% of the entire area, and C for slates with high severity damage covering more than 10% of the entire area. Fig. 2 illustrates the criteria for severity of damage of slates. 2.3. Sampling and analysis method For sampling and analysis, methods adopted by the US National Institute for Occupational Safety and Health (NIOSH), US Occupational Safety & Health Administration (OSHA), UK Health and Safety Executive (HSE) and Korea Occupational Safety & Health Agency (KOSHA) were compared and analyzed. Slates were sampled according to the NIOSH 7400 method, and analyzed according to the NOISH 7400a method and the OSHA ID-160 method (NIOSH, 1994; OSHA, 1997; HSE, 1998; KOSHA, 2006). The NIOSH 7400 method specifies that air is sampled at the height of 1500 mm above ground by using a mixed cellulose
91
ester membrane filter with a diameter of 25 mm and pores of 0.8 mm. For increased sample precision, flow rate of the pump and sampling time should be adjusted so that the density of the fiber should be 100e1300 pieces/mm2. For this purpose, sufficient amount of air (3000e10,000 L) should be sampled from relatively clean air, and air should be sampled for 8 h at the rate of 1e4 L/min from normal air. For emergencies with high concentration, air should be sampled for a short period of time at the rate of 7e16 L/min. It should be noted that dust should not cover more than 50% of the filter paper. For this study, the researchers sampled air at the rate of 14 L/min for 40 min per sample, assuming that damage of a slate would cause an emergency exposure to asbestos of high concentration and generate a large volume of dust. When sampling air, the researchers set the opening of the filter to face downward, and kept compensating the flow by using a pump. Asbestos samples can be analyzed with PCM, SEM or TEM methods. In this research, a combination of PCM and SEM was used. The PCM method provides fast analysis thanks to simple preprocessing, but does not distinguish accurately between asbestos fibers and non-asbestos fibers as it uses the Aspect Ratio to detect asbestos fibers. To complement defects the researchers mounted the Energy Dispersive X-ray (EDX) on the SEM to detect asbestos by analyzing particles shapes and characteristics of the elements, then, they used PCM to count asbestos fibers. The SEM used for analysis was TESCAN's MIRA FEG SEM (accelerating voltage: 15 kV, working distance: 21 mm. The magnification is determined considering the size of each particle and the resolution), and the EDX was IXRF Systems' XRF (live time 100 s). The PCM used for analysis was NIKON's NieU PH (with 400 magnification, fibers of 5 um or longer and the Aspect Ratio of over 3:1 were counted), and the scale was a Walton-Beckett graticule with the diameter of 100 mm (area: 0.00785 mm2). Fig. 3 illustrates a Walton-Beckett graticule. 3. Measurement of asbestos fiber generated from slates This section describes the methods of experiment and analysis to measure asbestos fiber generated from slate. It describes how to collect slates, how to classify slate based on the damage severity criteria, how to configure the experiment chamber, and how to sample asbestos fiber. This section also provides description on the SEM method to identify asbestos fiber, the PCM method to count asbestos fiber, and the NIOSH's formula to make a quantitative model. 3.1. Experiment method Before starting the study, the researchers collected slates used in typical buildings in Korea. After securing the buildings with slates, two researchers visited the sites, collected samples according to the
Fig. 4. Criteria for severity of damage of slates.
92
Y.-C. Kim et al. / Journal of Cleaner Production 86 (2015) 88e97
Fig. 5. Structure with slotted angle bars and configuration of equipment inside the chamber.
procedures specified in the Asbestos Safety Management Act of Korea (Ministry of Government Legislation, 2014) and carried the samples to the test facility. The researchers removed dust and other foreign substances from the slates, and dried the slates in the sun for an appropriate time. According to the severity of damage of slates as specified in paragraph 2.2, the researchers classified the slates into A (good), B (normal) and C (bad), and selected 3 slates from each grade, making 9 test samples in total. The researchers numbered the selected nine slates (A1, A2, A3; B1, B2, B3; C1, C2, C3), and measured their weight and area. Fig. 4 illustrates the criteria for severity of damage of slates. Then, to sample air for each grade of slate damage, the researchers free-dropped the slates from the roof height of typical buildings. To block external air and to secure accurate air samples, the experiment was conducted inside an experiment chamber. To simulate natural damage, a structure was installed with slotted angle bars inside the chamber. The slates slid down over the inclined plane of the angle bar structure and fell free at the edge of the inclined plane. Asbestos fiber was generated and emitted to the air as slates fell and were damaged. The chamber was designed in a hexagonal shape of 5000 mm (length) 2000 mm (width) 5000 mm (height) with a double-wall vinyl sheet inside the chamber for a complete seal. Fig. 5 illustrates the structure with
slotted angle bars and configuration of the equipment inside the chamber. In Korea, slates are widely used as roof material for single-story residential buildings. Therefore, the structure was assembled in reference to the typical residential building in Korea. The roof of the angle bar structure was designed to be inclined by 25 and 2500 mm high (Lee, 2002). The chamber had a hole for slates at the side so that slates could be supplied to the roof of the angle bar structure. On the opposite side of the chamber were a PCM asbestos sampler and an MCE filter to sample air containing asbestos fiber generated from damaged slates. Air was sampled at the height of 1500 mm from the ground according to the air sampling method of NIOSH 7400, and the MCE filter of 0.8 mm, ø25 mm was used. When damaged, slates generated a large amount of dust, as well as asbestos fibers. Too much dust collected caused error in the counting result. To make the analysis easier, the experiment was set with a flow of 14 L/min and a sampling time of 40 min. Each collected sample was 560 L of air. The researchers collected air from A1 slate, selected larger pieces of damaged slates, gave the sample numbers again, and repeated the aforementioned experiment procedure. The experiment was repeated until 10 air samples were acquired from each of the 9 slates; 90 air samples were made.
Fig. 6. SEM photo showing the shape of asbestos fiber and the EDX composition chart.
Y.-C. Kim et al. / Journal of Cleaner Production 86 (2015) 88e97
93
Table 2 Result of the experiment. Group A
Group B
Group C
Weight (kg) Area (m2) Density (f/mm2) Conc. (f/cc) Weight (kg) Area (m2) Density (f/mm2) Conc. (f/cc) Weight (kg) Area (m2) Density (f/mm2) Conc. (f/cc) 16.06 15.40 15.30 12.30 11.32 10.58 10.45 9.78 9.32 8.74 7.63 6.56 6.28 5.78 5.72 5.24 4.68 4.32 3.96 3.24 3.21 3.20 2.75 2.50 1.48 1.42 1.28 1.26 1.12 1.12
1.295 1.295 1.294 1.034 0.958 0.895 0.843 0.823 0.788 0.705 0.642 0.529 0.531 0.466 0.481 0.443 0.394 0.348 0.319 0.272 0.272 0.258 0.233 0.210 0.124 0.115 0.103 0.107 0.094 0.095
361.8 324.8 332.5 322.3 317.2 349.0 322.3 231.8 305.7 249.7 198.7 168.2 158.0 193.6 186.0 267.5 238.2 93.0 222.9 73.9 152.9 152.9 151.6 122.3 177.1 180.9 220.4 132.5 122.3 113.4
0.249 0.223 0.229 0.222 0.218 0.284 0.222 0.159 0.254 0.172 0.137 0.116 0.109 0.133 0.128 0.184 0.164 0.064 0.153 0.051 0.105 0.105 0.104 0.084 0.122 0.212 0.195 0.091 0.078 0.084
18.40 17.40 16.24 15.70 15.30 14.72 13.21 13.20 11.98 10.78 10.56 10.35 9.52 8.65 8.35 7.65 6.32 6.26 5.43 5.26 4.56 4.50 4.21 3.60 3.30 2.82 2.52 1.62 1.54 1.08
1.508 1.468 1.370 1.463 1.254 1.242 1.083 1.230 1.011 0.884 0.984 0.873 0.780 0.730 0.778 0.713 0.518 0.583 0.506 0.444 0.425 0.380 0.345 0.295 0.278 0.238 0.207 0.151 0.144 0.089
380.9 436.9 338.9 487.9 326.1 401.3 272.6 465.0 377.1 245.9 319.7 309.6 233.1 366.9 219.1 257.3 174.5 300.6 214.0 214.0 242.0 184.7 141.4 146.5 207.6 87.9 165.6 136.3 91.7 75.2
0.174 0.300 0.233 0.335 0.224 0.276 0.187 0.320 0.259 0.169 0.220 0.213 0.160 0.252 0.151 0.177 0.120 0.207 0.147 0.147 0.166 0.127 0.053 0.057 0.187 0.017 0.114 0.094 0.019 0.008
12.23 11.56 10.26 10.12 9.24 8.88 8.56 8.18 8.00 7.62 7.62 7.22 5.88 5.36 5.24 4.98 4.50 4.47 4.32 3.55 3.38 2.76 2.55 2.49 2.04 1.62 1.24 1.18 1.04 1.00
1.182 1.099 1.089 0.978 0.981 0.943 0.827 0.868 0.760 0.724 0.809 0.698 0.559 0.518 0.556 0.481 0.435 0.475 0.411 0.343 0.359 0.267 0.271 0.237 0.217 0.154 0.120 0.112 0.099 0.095
490.4 384.7 491.7 470.1 459.9 361.8 322.3 299.4 440.8 332.5 378.3 259.9 310.8 299.4 307.0 221.7 217.8 335.0 275.2 214.0 113.4 123.6 196.2 160.5 131.2 87.9 143.9 131.2 216.6 191.1
0.337 0.264 0.338 0.323 0.316 0.249 0.222 0.206 0.303 0.229 0.260 0.179 0.214 0.206 0.211 0.152 0.150 0.230 0.189 0.147 0.078 0.085 0.135 0.110 0.090 0.060 0.143 0.090 0.193 0.219
Conc.: Concentration.
3.2. Method of analysis To count the asbestos fibers generated from the damaged slate, the SEM method specified in OSHA ID-160 and the PCM method specified in NIOSH 7400a were used. OSHA recommends SEM to detect asbestos before PCM is applied for analysis. SEM can be applied for analysis of both the coefficient of asbestos fiber among the sampled air in the filter and the content of asbestos in the solid sample. SEM shows the chemical composition of the fiber material, enabling the researcher to determine whether the filtered particle is actually asbestos and the type of asbestos it is. PCM causes less damage to samples, provides relatively simple preprocessing, and enables fast analysis. Therefore, to raise the accuracy of analysis in this research, SEM was applied first to detect asbestos particles based on the chemical composition, and then, PCM to account asbestos fibers. To determine if the filtered fiber particle is asbestos by analyzing the physical and chemical characteristics, EDX was mounted on SEM to analyze 90 air samples. The shape of the fiber particles shown in the screen in the SEM observation was similar to the particles observed from the standard asbestos material. The EDX analysis showed that average proportions of elements was Mg
Table 3 Pearson correlation between sheet area and fiber concentration. Division
Pearson correlation
Sig.
All Group A Group B Group C
0.839** 0.845** 0.901** 0.900**
0.000 0.000 0.000 0.000
* : p < 0.05, **: p < 0.01. Sig.: significant probability.
(49.8%), Si (45.7%) and Fe (4.5%), which is similar to that of chrysotile. Fig. 6 demonstrates the SEM photo showing the shape of asbestos fibers and the EDX composition chart. After determining that the fiber particles had the chemical composition of chrysotile by using EDX mounted on SEM, PCM was used to count the asbestos fibers among the air samples. 1/4 of the MCE filter paper was cut and preprocessed with the acetonetriacetin solution. The particle was magnified by 400 with the Walton-Backett graticule mounted on PCM, and asbestos fibers were counted within the graticule with the diameter of 100 mm on the microscope. Fibers of 5 mm or longer with a 3:1 ratio between length and width were counted, and sufficient number of graticule fields were used so that at least 100 fibers should be counted. The numbers of asbestos fiber particles and the concentration in the air sample were estimated with the following formula (NIOSH, 1994):
Ns ¼
C¼
NA an
NA a V n 1000
Where, Ns: The total number of fiber particles C: Concentration of asbestos fiber in the air (f/cc) N: The number of fiber particles counted A: The area from which the valid sample is collected (385 mm2) a: The area of graticule in which fiber particles are counted (0.0078 mm2) V: The amount of air sampled (L) n: The total number of graticule fields in which fiber particles are counted
94
Y.-C. Kim et al. / Journal of Cleaner Production 86 (2015) 88e97
Table 4 Results of the linear regression analysis of the amount of asbestos fiber generated per unit area of damaged slate.
Intercept Sheet Area
Coefficient
Standard error
0.077 0.159
0.008 9.683** 0.011 14.491** F ¼ 210.002** (P e value ¼ 0.000) R2(adjR2) ¼ 0.701 Regression equation of asbestos fiber concentration C ¼ 0.077 þ 0.159 A
t
Significance probability 0.000 0.000 * : p < 0.05,**: p < 0.01 C: Fiber Concentration A: Area of Slate
Fig. 7. Amount of asbestos fiber generated per unit area of damaged slate.
4. Estimation of the amount of asbestos fiber generated from damaged slates Sample results extracted by NIOSH 7400a are substituted to the formula mentioned in Section 3.2 to develop a database of asbestos fiber concentration of each air sample. It is then classified by evaluating the grade of the severity of damage mentioned in Section 2.2 to perform general analysis and regression analysis. Pearson correlation analysis was used to find out the area of asbestos slate that had an impact on the concentration of asbestos fibers. Finally the researchers conducted linear regression analysis with SPSS 20.0. 4.1. Analysis of asbestos fibers per unit area of slate In this research, 3 slates were selected for each of the 3 grades, and the air was sampled for each of the 9 slates; a total 90 air samples were analyzed with SEM/PCM. The average density of the slate was grade A (12.02 kg/m2) > grade B (11.60 kg/m2) > grade C (10.03 kg/m2). Area of the slate was between 0.089 m2 and 1.508 m2. This result shows that the grades of the slates assessed based on the damage as specified in paragraph 2.2 were appropriate. The PCM analysis showed that the density of asbestos fibers collected with the MCE filter was between 73.9 f/mm2 and 491.7 f/ mm2 and the concentration of asbestos fibers was between 0.051 f/ cc and 0.338 f/cc. The indoor air quality standard (0.01 f/cc) established by the Ministry of Environment was exceeded in all the 90 samples. Therefore, it was found that slates, damaged as they fall from the roof of a typical building in Korea, scatter asbestos fiber particles at a much higher concentration than the standard value, regardless of the severity of damage. Table 2 provides the results of the experiment.
Based on the analysis of 90 air samples, the researchers estimated the amount of asbestos fibers generated per unit area of the damaged slate. First, SPSS was used to find out that the area of asbestos slate had an impact on the concentration of asbestos fiber. According to Pearson correlation analysis, the Pearson correlation coefficient, which indicates the degree of correlation, was 0.839. This means a positive linear relationship was strong and a significant value at 0.01 in significance level. In short, the area of an asbestos slate is a factor that affects the concentration of asbestos fiber. A Pearson correlation analysis performed to each grade classified by the severity of the damaged slate found a correlation coefficient statistics to stand at 0.845, 0.901 and 0.900; which are all significant at 0.01 in significance level. Results of the Pearson correlation analysis are in Table 3 below. The researchers conducted linear regression analysis with SPSS 20.0, with the area of the slate as independent variable and the concentration of asbestos fiber as dependent variable. Results of the analysis are listed in Table 4. The analysis showed that the P value 0.000 was appropriate at the significance level of 0.05, and the modified R2 value was 0.701. When a slate is damaged as it falls freely from the roof height of a building, the concentration of asbestos fiber grows by 0.159 f/cc as the area of the slate grows by 1 m2. Fig. 7 illustrates the result of the analysis. 4.2. Analysis of the amount of asbestos fiber generated per unit area of slate by grade The researchers calculated the amount of asbestos fiber generated per unit area of damaged slate by the grade of damage as deduced in paragraph 2.2. Based on the results of the analysis of 30 air samples for each of the 3 grades (A, B, C), the researchers conducted a linear regression analysis with SPSS 20.0, with the area of
Y.-C. Kim et al. / Journal of Cleaner Production 86 (2015) 88e97
95
Table 5 Results of the linear regression analysis of the amount of asbestos fiber generated per unit area of damaged slate (Group A). Group A
Coefficient
Standard error
Intercept Sheet Area
0.080 0.127
0.010 8.040** 0.015 8.350** F ¼ 69.724** (P e value ¼ 0.000) R2(adjR2) ¼ 0.703 Regression equation of asbestos fiber concentration CA ¼ 0.080 þ 0.127 AslateA
t
Significance probability 0.000 0.000 * : p < 0.05,**: p < 0.01 CA: Fiber Concentration (A Group) AslateA: Area of Slate (A Group)
Table 6 Results of the linear regression analysis of the amount of asbestos fiber generated per unit area of damaged slate (Group B). Group B
Coefficient
Standard error
Intercept Sheet Area
0.064 0.157
0.012 5.283** 0.014 10.992** F ¼ 120.833** (P e value ¼ 0.000) R2(adjR2) ¼ 0.805 Regression equation of asbestos fiber concentration CB ¼ 0.064 þ 0.157 AslateB
t
Significance probability 0.000 0.000 * : p < 0.05,**: p < 0.01 CB: Fiber Concentration (B Group) AslateB: Area of Slate (B Group)
Table 7 Results of the linear regression analysis of the amount of asbestos fiber generated per unit area of damaged slate (Group C). Group C
Coefficient
Standard error
Intercept Sheet Area
0.069 0.221
0.013 5.310** 0.020 10.943** F ¼ 119.746** (P e value ¼ 0.000) R2(adjR2) ¼ 0.804 Regression equation of asbestos fiber concentration CC ¼ 0.069 þ 0.221 AslateC
the slate as independent variable and the concentration of asbestos fiber as dependent variable. For grade A slates, the analysis showed that the P value 0.000 was appropriate at the significance level of 0.05, and the modified R2 value was 0.703. This shows that for grade A slates, the concentration of asbestos fiber grows by 0.127 f/cc as the damaged area grows by 1 m2. For grade B slates the analysis showed that the P value 0.000 was appropriate at the significance level of 0.05, and the modified R2 value was 0.805. This shows that for grade B slates, the concentration of asbestos fiber grows by 0.157f/cc as the damaged area grows by 1 m2. For grade C slates the analysis showed that the P value 0.000 was appropriate at the significance level of 0.05, and the modified R2 value was 0.804. This shows that for grade C slates the concentration of asbestos fiber grows by 0.221 f/cc as the damaged area grows by 1 m2. Tables 5e7 show the result of regression analysis for each grade, and the regression expression about the amount of asbestos fiber generated per unit area by grade. 4.3. Comparison of the amount of asbestos fiber generated from the slate between grades The amount of asbestos fiber generated from the damaged slate was compared between grades A, B and C. Fig. 8 illustrates the amount of asbestos fiber generated per unit area of the damaged slate by grade of damage. The result of the analysis in paragraph 4.2 shows that the concentration of asbestos fiber per unit area was 0.127 f/cc for grade A, 0.157 f/cc for grade B, and 0.221 f/cc for grade C. This shows that more asbestos fiber is generated as the slate is more damaged. When the amount of asbestos fiber generated by grade is compared with the analysis of paragraph 4.1 (average amount of asbestos fiber per unit area; 0.159 f/cc), grade C slates generate more asbestos fiber than the average by about 39.0%, and grade A generates less by about 20.1%.
t
Significance probability 0.000 0.000 * : p < 0.05,**: p < 0.01 CC: Fiber Concentration (C Group) AslateC: Area of Slate (C Group)
5. Conclusions Use of asbestos is banned or restricted around the world but there still exists a large amount of asbestos slates. An effective maintenance, disintegration and removal of asbestos slates during normal times are critical. Of equal importance is quantifying the amount of asbestos generated during disaster situations like earthquakes as a way to develop solutions to keep asbestos from scattering. While many preceding research focused on the harmfulness of asbestos and asbestos release from slates, so far no research has looked into asbestos fiber generated from a damaged slate. This raises the need to quantify the amount of asbestos fiber generated when a slate is damaged. This research estimated the amount of asbestos fiber generated per unit area of damaged slates by severity of damage. Used slates were collected from buildings and divided into three grades depending on the severity of damage. To simulate natural damage, the researchers free-dropped slates from the roof height of a typical building and sampled the air surrounding it. The asbestos fiber particles were counted with SEM/PCM in the air samples and classified into three grades. Based on the results, the researchers conducted a regression analysis with SPSS 20.0, and developed a model for estimating the amount of asbestos fiber generated per unit area of damaged slate. SEM/PCM analysis of air samples showed that damaged slates emitted chrysotile, and the content of asbestos fiber emitted to the air was between 0.051 f/cc and 0.338 f/cc. It was found that the indoor air quality standard of 0.01 f/cc established by the Ministry of Environment was exceeded in all 90 samples. This shows that damaged slates are very dangerous as they emit a large amount of asbestos fibers. Linear regression analysis showed that damaged slates generate asbestos fibers of 0.159 f/cc per unit area. Tables 5e7 show the regression expressions for the amount of asbestos fiber generated by grade. The generation of asbestos fiber per unit area
96
Y.-C. Kim et al. / Journal of Cleaner Production 86 (2015) 88e97
Fig. 8. Amount of asbestos fiber generated per unit area of damaged slate by grade of damage.
was 0.127 f/cc for grade A, 0.157 f/cc for grade B, and 0.221 f/cc for grade C, which shows that more asbestos fiber is generated as damage of the slate worsens. The researchers developed the asbestos fiber calculation formula for each grade. When the amount of asbestos fiber generated by grade is compared with the average amount of asbestos fiber per unit area, grade C slate generates more asbestos fiber than the average by about 39.0%, and grade A generates less by about 20.1%. This research revealed that asbestos materials used in buildings are far more hazardous when damaged than when they are not. It was found that, the greater the damage, more asbestos fiber would scatter. The estimation model offered in this study can quantify the amount of asbestos fiber based on the building area by estimating the amount of asbestos fiber scattered due to the slate's severity of damage. This estimation model is expected to contribute to the quantification of asbestos generated due to disasters, including earthquakes, by building and by area, which were not been possible before due to the lack of relevant studies. A formula to determine the amount of asbestos fiber from this research is needed at a time when research to control asbestos is increasingly becoming a critical issue around the world. It is very important to quantify asbestos fiber emitted due to disasters and formulate measures to prevent scattering as to maintain, dismantle and remove asbestos slates effectively in normal times. In this sense, the asbestos fiber estimating formula suggested in this study is deemed important. The result of this study may be utilized as basic data to estimate the influence of the slate present in buildings on the environment and human body, and as a guideline to prevent damages to asbestos and to handle asbestos in case of disaster. In future research, a more indepth research evaluating ambient environment, which accounts for air movement, and risk to the human body based on findings of this research to develop a simulation on the spread of asbestos fiber will hopefully support the efficient use of the simulation in disaster situations.
Acknowledgments This research was supported by a grant (12 High-tech Urban C20) from High-tech Urban Development Program funded by Ministry of Land, Infrastructure and Transport of Korea government (12 High-tech Urban C20).
References AHERA (Asbestos Hazard Emergency Response Act), 2003, 1987. Part 763, Subpart E Asbestos-containing Materials in Schools. EPA, 52 CFR Part 763. http://www2. epa.gov/sites/production/files/documents/2003pt763_0.pdf. ASTM (American Society for Testing and Materials), 2014. Standard Practice for Comprehensive Building Asbestos Surveys. DOI.10.1520/E2356-10. http://www. astm.org/Standards/E2356.htm. Bae, Jae-Min, Kim, Young-Chan, Hong, Won-Hwa, 2013. The development of estimation model for calculated CO2 emissions from dismantling and demolition of asbestos containing materials. Archit. Inst. Korea 29 (10), 267e274. Becklake, M.R., 1976. Asbestos-related diseases of the lung and other organs: their epidemiology and implications for clinical practice. Am. Rev. Respir. Dis. 1, 187e227. Beddoe, R.E., Dorner, H.W., 2005. Modelling acid attack on concrete: part I. The essential mechanisms. Cem. Concr. Res. 12, 2333e2339. Bornemann, P., Hildebrandt, U., 1986. On the problem of environmental pollution by weathering products of asbestos cement. Staub Reinhalt. Luft 487e489. Choi, Jung-Keun, Paek, Do-Myung, Paik, Nam-Won, 1998. The production, the use, the number of workers and exposure level of asbestos in Korea. Korean Ind. Hyg. Assoc. J. 8 (2), 242e253. Deng, Q., Lan, Y., Wang, M., 2009. 30 years follow up study: the dose-response relationship of asbestos exposure and asbestosis. Mod. Prev. Med. 36 (11), 2027e2029. Doll, R., Peto, J., Britain, G., 1985. Effects on Health of Exposure to Asbestos. Health & Safety Commission. Dyczek, J., 2006. Surface of asbestos-cement (AC) roof sheets and assessment of the risk of asbestos release. In: Conference Presentation Krakow, Poland, September 28e29. Ernst, W., 2012. Overview of naturally occurring earth materials and human health concerns. J. Asian Earth Sci. 108e126. Favero-Longo, S., Castelli, D., Fubini, B., Piervittori, R., 2009. Lichens on asbestosecement roofs: bioweathering and biocovering effects. J. Hazard. Mater. 2, 1300e1308. Hourihane, D.O., 1964. The pathology of mesotheliomata and an analysis of their association with asbestos exposure. Thorax 3, 268e278. Health and Safety Executive (HSE), 1998. MDHS 87 Fibres in Air: Guidance on the Discrimination Between Fibre Types in Samples of Airborne Dust on Filters Using Microscopy. http://www.hse.gov.uk/pubns/mdhs/pdfs/mdhs87.pdf. HSE (Health and Safety Executive), 2002. A Comprehensive Guide to Managing Asbestos in Premises. ISBN 978 0 7176 2381 5/HSG227. http://www.hse.gov.uk/ pubns/priced/hsg227.pdf. Jinhui, L., Qingyin, D., Keli, Y., Lili, L., 15 October 2014. Asbestos and asbestos waste management in the Asian-Pacific region: trends, challenges and solutions. J. Clean. Prod. 81, 218e226. Kamp, D.W., 2009. Asbestos-induced lung diseases: an update. Transl. Res. 4, 143e152. Kane, A.B., Boffetta, P., Saracci, R., Wilbourn, J., 1996. Mechanisms of Fibre Carcinogenesis. International Agency for Research on Cancer. Kazan-Allen, L., 2005. Asbestos and mesothelioma: worldwide trends. Lung Cancer S3eS8. Kim, Ji-Yeong, Lee, Song-Kwon, Lee, Jeong-Hee, Lim, Mu-Heok, Kang, Sung-Wook, Phee, Young-Gyu, 2009. Study on the factors affecting asbestos exposure level from asbestos abatement in building. J. Korean Soc. Occup. Environ. Hyg. 19 (1), 8e15.
Y.-C. Kim et al. / Journal of Cleaner Production 86 (2015) 88e97 Kim, Hyun-Wook, Park, Gye-Young, Han, Jin-Gu, Han, Young-Sun, Hwang, Bum-Gu, Lee, Jun-Hyuk, 2010. Releasing of asbestos fibers from the weathered asbestos cement slate roofing. J. Korean Soc. Occup. Environ. Hyg. 20 (2), 88e93. Kim, Y., Son, B., Park, W., Hong, W., 2011a. A study on the distribution of the asbestos cement slates and calculation of disposal cost in the rural area. Archit. Res. 2, 31e40. Kim, Young-Chan, Son, Byeung-Hun, Kim, Hye-Mi, Hong, Won-Hwa, 2011b. A study on the distribution maps for asbestos cement slates using GIS. J. Korea Inst. Ecol. Archit. Environ. 11 (3), 57e62. Kim, Young-Chan, Hong, Won-Hwa, Zhang, Yuan-Long, 2014. A study on the efficient method of information offering through survey of asbestos awareness of deteriorated apartment facility managers. Archit. Inst. Korea 30 (1), 217e224. Korea Occupational Safety and Health Agency, 2006. KOSHA CODE A-1-2004. Lee, Seung-Ho, 2002. The characteristics of folk houses at heavy snowfall regions in Korea. J. Korean Geogr. Soc. 37 (1), 1e14. Linton, A., Vardy, J., Clarke, S., van Zandwijk, N., 2012. The ticking time-bomb of asbestos: its insidious role in the development of malignant mesothelioma. Crit. Rev. Oncol. 2, 200e212. Meyer, E., 1986. Examination of the importance of weathering asbestos cement planes for concentration of asbestos fibers in the environment. Staub Reinhalt. Luft 482e483. Ministry of Government Legislation, 2014. Asbestos Safety Management Act. http:// www.law.go.kr/%EB%B2%95%EB%A0%B9/%EC%84%9D%EB%A9%B4%EC%95%88% EC%A0%84%EA%B4%80%EB%A6%AC%EB%B2%95. Nicholson, W.J., 2001. The carcinogenicity of chrysotile asbestosea review. Ind. Health 2, 57e64. National Institute for Occupational Safety and Health (NIOSH), 1994. Asbestos and Other Fibers by PCM: 7400, fourth ed. In: NIOSH Manual of Analytical Methods
97
(NMAM) S. Department of HHS http://www.cdc.gov/niosh/docs/2003-154/pdfs/ 7400.pdf. Occupational Safety and Health Administration (OSHA), 1997. OSHA ID-160, Asbestos in Air. OSHA. https://www.osha.gov/dts/sltc/methods/inorganic/ id160/id160.pdf. Pastuszka, J.S., 2009. Emission of airborne fibers from mechanically impacted asbestos-cement sheets and concentration of fibrous aerosol in the home environment in Upper Silesia, Poland. J. Hazard. Mater. 162 (2), 1171e1177. Selikoff, I.J., Churg, J., Hammond, E.C., 1964. Asbestos exposure and neoplasia. JAMA 1, 142e146. Sim, M.R., 2013. A worldwide ban on asbestos production and use: some recent progress, but more still to be done. Occup. Environ. Med. 1, 1e2. Spurny, K., 1989. On the release of asbestos fibers from weathered and corroded asbestos cement products. Environ. Res. 1, 100e116. The Ministry of Environment of Korea government, 2009. Asbestos Management Catalogue. http://www.me.go.kr/. Turci, F., Favero-Longo, S.E., Tomatis, M., Martra, G., Castelli, D., Piervittori, R., Fubini, B., 2007. A biomimetic approach to the chemical inactivation of chrysotile fibres by lichen metabolites. Chem. A Eur. J. 14, 4081e4093. U.S. EPA (Environmental Protection Agency), 2003. Managing Asbestos in Place: a Building Owner's Guide to Operations and Maintenance Programs for Asbestoscontaining Materials (Green Book). Office of Pesticides and Toxic Substances, Washington, DC. http://nepis.epa.gov/Exe/ZyPDF.cgi/20011E4D.PDF? Dockey¼20011E4D.PDF. Virta, R.L., 2006. Worldwide Asbestos Supply and Consumption Trends from 1900 Through 2003. US Geological Survey Reston, VA. Wagner, J., 1965. The sequelae of exposure to asbestos dust. Ann. N. Y. Acad. Sci. 1, 691e695.
Contents lists available at ScienceDirect
Journal of Cleaner Production journal homepage: www.elsevier.com/locate/jclepro
Development of a model to calculate asbestos fiber from damaged asbestos slates depending on the degree of damage Young-Chan Kim, Won-Hwa Hong*, Yuan-Long Zhang School of Architecture, Civil, Environmental and Energy Engineering, Kyungpook National University, 80 Daehak-ro, Buk-gu, Daegu 702-701, Republic of Korea
a r t i c l e i n f o
a b s t r a c t
Article history: Received 9 July 2014 Received in revised form 28 August 2014 Accepted 28 August 2014 Available online 6 September 2014
It is very important to quantify asbestos fiber emitted due to disasters and formulate measures to prevent scattering as to maintain, dismantle and remove asbestos slates effectively in normal times. The researchers developed a model to estimate the amount of asbestos fiber generated depending on the damage of slates, focusing on natural damages. The researchers collected used slates from typical buildings and classified them into 3 grades depending on the severity of damage. The researchers freedropped slates from the roof height of a typical building, and sampled the air surrounding it. With the air sampled for each of the 3 grades of damage, the researchers counted the asbestos fibers by using the Scanning Electron Microscopy/Phase Contrast Microscopy. By conducting a regression analysis with SPSS 20.0 based on data acquired, the researchers deduced a formula. The linear regression analysis showed that ‘damaged slate generates asbestos fiber per unit area ¼ 0.077 þ 0.159 area of slate (coefficient of determination: 70.1%)’. Generation of asbestos fiber per unit area was 0.127 f/cc (Good), 0.157 f/cc (Normal), and 0.221 f/cc (Bad). The estimation model offered in this study can quantify the amount of asbestos fiber based on the building area by estimating the amount of asbestos fiber scattered due to the slate's severity of damage. This estimation model is expected to contribute to the quantification of asbestos generated due to disasters, including earthquakes, by building and by area, which were not been possible before due to the lack of relevant studies. © 2014 Elsevier Ltd. All rights reserved.
Keywords: Asbestos Damaged asbestos cement slate Asbestos fiber Degree of damage
1. Introduction Asbestos has been used since ancient times due to its high affordability and desirable physical properties such as resistance to heat and fire as well as its antiseptic and insulating properties. With the utilization of steam engines since the Industrial Revolution, asbestos consumption increased rapidly (Becklake, 1976; Sim, M.R., 2013). Global consumption of asbestos started to increase rapidly in the 1940s and reached its peak in the 1980s (Virta, 2006). Exposure to asbestos, however, is known to cause incurable diseases after a latent period of 20e50 years such as pulmonary asbestosis, lung cancer with unfavorable prognosis and malignant mesothelioma (Doll et al., 1985; Hourihane, 1964; Linton A. et al., 2012). Harmfulness of asbestos has long been a subject of research (Wagner, 1965; Kamp, D.W., 2009). Due to this asbestos use has been reduced gradually since the 1970s and some countries have been establishing related laws to prohibit or limit development of
* Corresponding author. Tel.: þ82 53 950 5597; fax: þ82 53 950 6590. E-mail addresses: [email protected], [email protected] (W.-H. Hong). http://dx.doi.org/10.1016/j.jclepro.2014.08.092 0959-6526/© 2014 Elsevier Ltd. All rights reserved.
asbestos since the early 1990s (Kane et al., 1996; Nicholson, 2001; Selikoff et al., 1964; Deng et al., 2009). Nevertheless, asbestos is still used in large quantities in many countries as building material especially in the form of asbestos cement (Kazan-Allen, 2005; Jinhui et al., 2014). Asbestos-caused diseases generally occur through inhalation of asbestos fiber (Kane et al., 1996; Ernst, W., 2012). Asbestos cement is not deemed harmful to health since the asbestos fibers are strongly bound by the cement (U.S. EPA, 2003). However, it caused problems when asbestos fibers are lost and emitted to their surroundings through years of weathering (Bornemann and Hildebrandt, 1986). Use of asbestos is prohibited or limited in most countries but is increasing in Latin America, Russia and Asia (Kazan-Allen, 2005). Unlike the increasing use of asbestos in Asia, Korea started to import asbestos in the 1960s, it recorded its peak use in 1992 by about 95 thousand tons but has shown a consistent decrease since then (Kim et al., 2009). After inserting asbestos in the list of harmful substances that require permission to be used with the revision of the Occupational Safety and Health Act in 1990, Korea banned the use of asbestos completely with the revision of an enforcement decree of the act in 2009 (Korea ME, 2009). In addition, Korea has
Y.-C. Kim et al. / Journal of Cleaner Production 86 (2015) 88e97
89
Fig. 1. Process flow of this research.
established the Asbestos Injury Relief Act and the Asbestos Safety Management Act to regulate compensation for damage caused by asbestos and to control the safety of used asbestos (Kim et al., 2014). In Korea, about 96% of the imported asbestos in the 1970s was used in slates, a type of asbestos cement, and it was used about 82% of the time in the 1990s (J.K. Choi et al., 1998). In Korea, slates were manufactured in a mixture of about 90% cement and 10% chrysotile. All the slates produced were asbestos cement slates (Kim et al., 2010). Calcium hydroxide, a component of cement, is watersoluble (Beddoe and Dorner, 2005). As time passes, asbestos fibers contained in the slate are emitted into the surroundings (Bornemann and Hildebrandt, 1986) and solubility grows in acid rain (Dyczek, 2006). The high concentration of asbestos measured in the air around buildings with deteriorated slate roofs has a harmful influence on health (Spurny, 1989). About 80% of asbestos emitted from corroded surfaces is reported to be washed away by rain water with about 20% being emitted into the air, but the exact ratio has not been verified (Meyer, 1986). Recent research reported that moss can protect the surface of the slate as the metabolite of moss converts the chemical component of asbestos fibers into a harmless substance (Favero-Longo et al., 2009; Turci et al., 2007) however this is not an ultimate solution. Considering the global trend of banning asbestos, the situation in Korea, and the hazards of slates, the solution against asbestos-related problems in Korea should be effective maintenance, safe removal and safe disposal (Kim et al., 2011a,b). Having reviewed the studies conducted in Korea and overseas, the researchers could find a number of research on the health hazard of asbestos and emission of asbestos from slates, but could not find any research on the amount of asbestos fiber generated from damaged slates. The damage of a given slate is divided into artificial and natural damage (e.g., earthquake). This research focuses on natural damage as it is designed for application in disaster situations including earthquakes. Before starting the research, slates used in typical buildings were collected and classified into 3 grades depending on the severity of damage. To simulate natural damage, the researchers free-dropped slates from the roof height of a typical building and sampled the air surrounding it. The
researchers then counted the asbestos fibers in the sampled air depending on the severity of damage, and deduced a regression equation for the amount of asbestos fiber by the area of the slate. This study estimates the amount of asbestos fiber generated due to damage of slates by building and by district, and is expected to be widely used in policy decisions such as measures to prevent asbestos damages and development of guidelines. To deduce a regression equation for the amount of asbestos generated by the severity of damage of slates, the researchers collected used slates from typical buildings in Korea. To establish the severity of damage, the researchers compared and analyzed various methods to assess potential scattering and exposure of asbestos-containing materials such as: the US Asbestos Hazard Emergency Response Act (AHERA, 1987), the American Society for Testing and Materials (ASTM, 2014), and the UK Health & Safety Executive (HSE, 2002). As a result of the comparative analysis, the researchers established the appropriate criteria for assessment of the damage of slates, established 3 grades (A, B and C) for damage of the collected slates, selected 3 slates from each grade, and measured the area and weight of each slate. To conduct the experiment, the researchers constructed a double-walled vinylsealed chamber and free-dropped the slates of each grade from the roof height of a typical building. The asbestos fibers generated from the damaged slate in the chamber were sampled with the asbestos samplers and mixed cellulose ester membrane filter (MCE), counted with the Scanning Electron Microscopy (SEM)/Phase Contrast Microscopy (PCM), and the results entered into a database. By conducting a regression analysis with SPSS 20.0 based on the database, the researchers deduced the formula to calculate the amount of asbestos fiber depending on the severity of damage of the slate. Fig. 1 illustrates the process flow of this research. 2. Method and application of theories In order to raise understanding of this study, this section describes types/characteristics and distribution of asbestoscontaining materials used in Korea, and provides the overview of a slate. To establish severity of damage, the researchers compared
90
Y.-C. Kim et al. / Journal of Cleaner Production 86 (2015) 88e97
Table 1 Specifications of slates specified in the Korean Standard. Type Small Corrugation
Large Corrugation
No. No. No. No. No. No.
6 7 8 6 7 8
Width (mm)
Length (mm)
Thickness (mm)
Average weight (kg/sheet)
Number of corrugation
Depth of corrugation (mm)
Asbestos content (%)
720 720 720 960 960 960
1800 2100 2400 1800 2100 2400
6.3
14 16 18 18 21 24
11.5
15
10
7.5
35
and analyzed various methods to assess potential scattering and exposure of asbestos-containing materials, such as the US Asbestos Hazard Emergency Response Act (AHERA), the American Society for Testing and Materials (ASTM), and the UK Health & Safety Executive (HSE), and established the criteria to assess damage of slates. Lastly, this section describes sampling and analysis methods used in this study.
2.1. Overview of a slate The asbestos-containing materials for typical buildings in Korea include slates (outdoor roof material), tex (outdoor ceiling finishing material), bamlite (indoor wall finishing material), gasket (facility joint), asbestos cloth (fire-fighting sheath), etc. (Bae et al., 2013). Slates have been used as a roof material for 1,260,000 buildings (18.09%) in Korea. When compared with indoor asbestoscontaining materials, slates are more easily dropped and damaged by external factors (vibration, wind, etc.). Damaged slates can generate asbestos fiber, which has a harmful influence to the environment and health. Therefore, research on scattering asbestos from slates has been conducted continuously. Firstly, Bornemann and Hildebrandt (1986) reported that an older slate roof which has long been installed or a slate roof with a damaged surface emits asbestos fibers to the air at an annual average rate of 3 g/m2 (Bornemann and Hildebrandt, 1986). Secondly, Spurny (1989) reported that asbestos cement slates would have corroded surfaces due to the weather changes at the rate of about 0.01e0.024 mm a year (Spurny, 1989). Thirdly, Pastuszka (2009) reported that asbestos cement slates would generate asbestos fibers at the rate of 2.7e6.9 103 F/(m2 J) when impacted (Pastuszka, 2009). So far,
however, there has been no research on the amount of asbestos fiber generated from damaged slates. Therefore, this research is aimed at the deduction of a regression equation to estimate the amount of asbestos fiber generated from a damaged slate out of asbestos-containing materials. In Korea, slates were manufactured by two companies, BYUCKSAN and KCC, in accordance with Korean standards. Therefore, all slate products used in Korea contained asbestos (Kim et al., 2011b). Slates are classified into small-corrugated slates and largecorrugated slates depending on the size of corrugation. Largecorrugated slates are normally used in plants and stalls, while small-corrugated slates are normally used in general buildings. Therefore, this research was conducted with small-corrugated slates produced in Korea. Table 1 shows specifications of the slates as per Korean standards (Kim et al., 2011a,b).
2.2. Slate damage assessment method Methods of assessing damage of asbestos-containing materials include: AHERA and ASTM of USA, and HSE of UK. The AHERA method divides the damage of asbestos-containing materials into serious damage, medium damage and good depending on the severity of damage. This method provides a simple assessment system and enables a systematic integration. The ASTM method divides the state of the samples into good, normal and bad depending on the severity of damage. This method provides scoring of each risk factor, and enables building managers and owners to actively take measures based on their own judgment. The HSE method divides the state of asbestos-containing materials into good, slight damage, medium damage and serious damage. This
Fig. 2. Criteria for severity of damage of slates.
Y.-C. Kim et al. / Journal of Cleaner Production 86 (2015) 88e97
Fig. 3. Walton-Beckett graticule.
method sets the priority order of management, assessing dustability and hazard, respectively. For this research, the researchers compared and analyzed the above three methods and deduced the appropriate grades for the damage of slates. The states of the slates are divided into grades A (good), B (normal) and C (bad) depending on the severity of damage: grade A is given for slates with no or minimum damage, B for slates with medium severity damage covering not more than 10% of the entire area, and C for slates with high severity damage covering more than 10% of the entire area. Fig. 2 illustrates the criteria for severity of damage of slates. 2.3. Sampling and analysis method For sampling and analysis, methods adopted by the US National Institute for Occupational Safety and Health (NIOSH), US Occupational Safety & Health Administration (OSHA), UK Health and Safety Executive (HSE) and Korea Occupational Safety & Health Agency (KOSHA) were compared and analyzed. Slates were sampled according to the NIOSH 7400 method, and analyzed according to the NOISH 7400a method and the OSHA ID-160 method (NIOSH, 1994; OSHA, 1997; HSE, 1998; KOSHA, 2006). The NIOSH 7400 method specifies that air is sampled at the height of 1500 mm above ground by using a mixed cellulose
91
ester membrane filter with a diameter of 25 mm and pores of 0.8 mm. For increased sample precision, flow rate of the pump and sampling time should be adjusted so that the density of the fiber should be 100e1300 pieces/mm2. For this purpose, sufficient amount of air (3000e10,000 L) should be sampled from relatively clean air, and air should be sampled for 8 h at the rate of 1e4 L/min from normal air. For emergencies with high concentration, air should be sampled for a short period of time at the rate of 7e16 L/min. It should be noted that dust should not cover more than 50% of the filter paper. For this study, the researchers sampled air at the rate of 14 L/min for 40 min per sample, assuming that damage of a slate would cause an emergency exposure to asbestos of high concentration and generate a large volume of dust. When sampling air, the researchers set the opening of the filter to face downward, and kept compensating the flow by using a pump. Asbestos samples can be analyzed with PCM, SEM or TEM methods. In this research, a combination of PCM and SEM was used. The PCM method provides fast analysis thanks to simple preprocessing, but does not distinguish accurately between asbestos fibers and non-asbestos fibers as it uses the Aspect Ratio to detect asbestos fibers. To complement defects the researchers mounted the Energy Dispersive X-ray (EDX) on the SEM to detect asbestos by analyzing particles shapes and characteristics of the elements, then, they used PCM to count asbestos fibers. The SEM used for analysis was TESCAN's MIRA FEG SEM (accelerating voltage: 15 kV, working distance: 21 mm. The magnification is determined considering the size of each particle and the resolution), and the EDX was IXRF Systems' XRF (live time 100 s). The PCM used for analysis was NIKON's NieU PH (with 400 magnification, fibers of 5 um or longer and the Aspect Ratio of over 3:1 were counted), and the scale was a Walton-Beckett graticule with the diameter of 100 mm (area: 0.00785 mm2). Fig. 3 illustrates a Walton-Beckett graticule. 3. Measurement of asbestos fiber generated from slates This section describes the methods of experiment and analysis to measure asbestos fiber generated from slate. It describes how to collect slates, how to classify slate based on the damage severity criteria, how to configure the experiment chamber, and how to sample asbestos fiber. This section also provides description on the SEM method to identify asbestos fiber, the PCM method to count asbestos fiber, and the NIOSH's formula to make a quantitative model. 3.1. Experiment method Before starting the study, the researchers collected slates used in typical buildings in Korea. After securing the buildings with slates, two researchers visited the sites, collected samples according to the
Fig. 4. Criteria for severity of damage of slates.
92
Y.-C. Kim et al. / Journal of Cleaner Production 86 (2015) 88e97
Fig. 5. Structure with slotted angle bars and configuration of equipment inside the chamber.
procedures specified in the Asbestos Safety Management Act of Korea (Ministry of Government Legislation, 2014) and carried the samples to the test facility. The researchers removed dust and other foreign substances from the slates, and dried the slates in the sun for an appropriate time. According to the severity of damage of slates as specified in paragraph 2.2, the researchers classified the slates into A (good), B (normal) and C (bad), and selected 3 slates from each grade, making 9 test samples in total. The researchers numbered the selected nine slates (A1, A2, A3; B1, B2, B3; C1, C2, C3), and measured their weight and area. Fig. 4 illustrates the criteria for severity of damage of slates. Then, to sample air for each grade of slate damage, the researchers free-dropped the slates from the roof height of typical buildings. To block external air and to secure accurate air samples, the experiment was conducted inside an experiment chamber. To simulate natural damage, a structure was installed with slotted angle bars inside the chamber. The slates slid down over the inclined plane of the angle bar structure and fell free at the edge of the inclined plane. Asbestos fiber was generated and emitted to the air as slates fell and were damaged. The chamber was designed in a hexagonal shape of 5000 mm (length) 2000 mm (width) 5000 mm (height) with a double-wall vinyl sheet inside the chamber for a complete seal. Fig. 5 illustrates the structure with
slotted angle bars and configuration of the equipment inside the chamber. In Korea, slates are widely used as roof material for single-story residential buildings. Therefore, the structure was assembled in reference to the typical residential building in Korea. The roof of the angle bar structure was designed to be inclined by 25 and 2500 mm high (Lee, 2002). The chamber had a hole for slates at the side so that slates could be supplied to the roof of the angle bar structure. On the opposite side of the chamber were a PCM asbestos sampler and an MCE filter to sample air containing asbestos fiber generated from damaged slates. Air was sampled at the height of 1500 mm from the ground according to the air sampling method of NIOSH 7400, and the MCE filter of 0.8 mm, ø25 mm was used. When damaged, slates generated a large amount of dust, as well as asbestos fibers. Too much dust collected caused error in the counting result. To make the analysis easier, the experiment was set with a flow of 14 L/min and a sampling time of 40 min. Each collected sample was 560 L of air. The researchers collected air from A1 slate, selected larger pieces of damaged slates, gave the sample numbers again, and repeated the aforementioned experiment procedure. The experiment was repeated until 10 air samples were acquired from each of the 9 slates; 90 air samples were made.
Fig. 6. SEM photo showing the shape of asbestos fiber and the EDX composition chart.
Y.-C. Kim et al. / Journal of Cleaner Production 86 (2015) 88e97
93
Table 2 Result of the experiment. Group A
Group B
Group C
Weight (kg) Area (m2) Density (f/mm2) Conc. (f/cc) Weight (kg) Area (m2) Density (f/mm2) Conc. (f/cc) Weight (kg) Area (m2) Density (f/mm2) Conc. (f/cc) 16.06 15.40 15.30 12.30 11.32 10.58 10.45 9.78 9.32 8.74 7.63 6.56 6.28 5.78 5.72 5.24 4.68 4.32 3.96 3.24 3.21 3.20 2.75 2.50 1.48 1.42 1.28 1.26 1.12 1.12
1.295 1.295 1.294 1.034 0.958 0.895 0.843 0.823 0.788 0.705 0.642 0.529 0.531 0.466 0.481 0.443 0.394 0.348 0.319 0.272 0.272 0.258 0.233 0.210 0.124 0.115 0.103 0.107 0.094 0.095
361.8 324.8 332.5 322.3 317.2 349.0 322.3 231.8 305.7 249.7 198.7 168.2 158.0 193.6 186.0 267.5 238.2 93.0 222.9 73.9 152.9 152.9 151.6 122.3 177.1 180.9 220.4 132.5 122.3 113.4
0.249 0.223 0.229 0.222 0.218 0.284 0.222 0.159 0.254 0.172 0.137 0.116 0.109 0.133 0.128 0.184 0.164 0.064 0.153 0.051 0.105 0.105 0.104 0.084 0.122 0.212 0.195 0.091 0.078 0.084
18.40 17.40 16.24 15.70 15.30 14.72 13.21 13.20 11.98 10.78 10.56 10.35 9.52 8.65 8.35 7.65 6.32 6.26 5.43 5.26 4.56 4.50 4.21 3.60 3.30 2.82 2.52 1.62 1.54 1.08
1.508 1.468 1.370 1.463 1.254 1.242 1.083 1.230 1.011 0.884 0.984 0.873 0.780 0.730 0.778 0.713 0.518 0.583 0.506 0.444 0.425 0.380 0.345 0.295 0.278 0.238 0.207 0.151 0.144 0.089
380.9 436.9 338.9 487.9 326.1 401.3 272.6 465.0 377.1 245.9 319.7 309.6 233.1 366.9 219.1 257.3 174.5 300.6 214.0 214.0 242.0 184.7 141.4 146.5 207.6 87.9 165.6 136.3 91.7 75.2
0.174 0.300 0.233 0.335 0.224 0.276 0.187 0.320 0.259 0.169 0.220 0.213 0.160 0.252 0.151 0.177 0.120 0.207 0.147 0.147 0.166 0.127 0.053 0.057 0.187 0.017 0.114 0.094 0.019 0.008
12.23 11.56 10.26 10.12 9.24 8.88 8.56 8.18 8.00 7.62 7.62 7.22 5.88 5.36 5.24 4.98 4.50 4.47 4.32 3.55 3.38 2.76 2.55 2.49 2.04 1.62 1.24 1.18 1.04 1.00
1.182 1.099 1.089 0.978 0.981 0.943 0.827 0.868 0.760 0.724 0.809 0.698 0.559 0.518 0.556 0.481 0.435 0.475 0.411 0.343 0.359 0.267 0.271 0.237 0.217 0.154 0.120 0.112 0.099 0.095
490.4 384.7 491.7 470.1 459.9 361.8 322.3 299.4 440.8 332.5 378.3 259.9 310.8 299.4 307.0 221.7 217.8 335.0 275.2 214.0 113.4 123.6 196.2 160.5 131.2 87.9 143.9 131.2 216.6 191.1
0.337 0.264 0.338 0.323 0.316 0.249 0.222 0.206 0.303 0.229 0.260 0.179 0.214 0.206 0.211 0.152 0.150 0.230 0.189 0.147 0.078 0.085 0.135 0.110 0.090 0.060 0.143 0.090 0.193 0.219
Conc.: Concentration.
3.2. Method of analysis To count the asbestos fibers generated from the damaged slate, the SEM method specified in OSHA ID-160 and the PCM method specified in NIOSH 7400a were used. OSHA recommends SEM to detect asbestos before PCM is applied for analysis. SEM can be applied for analysis of both the coefficient of asbestos fiber among the sampled air in the filter and the content of asbestos in the solid sample. SEM shows the chemical composition of the fiber material, enabling the researcher to determine whether the filtered particle is actually asbestos and the type of asbestos it is. PCM causes less damage to samples, provides relatively simple preprocessing, and enables fast analysis. Therefore, to raise the accuracy of analysis in this research, SEM was applied first to detect asbestos particles based on the chemical composition, and then, PCM to account asbestos fibers. To determine if the filtered fiber particle is asbestos by analyzing the physical and chemical characteristics, EDX was mounted on SEM to analyze 90 air samples. The shape of the fiber particles shown in the screen in the SEM observation was similar to the particles observed from the standard asbestos material. The EDX analysis showed that average proportions of elements was Mg
Table 3 Pearson correlation between sheet area and fiber concentration. Division
Pearson correlation
Sig.
All Group A Group B Group C
0.839** 0.845** 0.901** 0.900**
0.000 0.000 0.000 0.000
* : p < 0.05, **: p < 0.01. Sig.: significant probability.
(49.8%), Si (45.7%) and Fe (4.5%), which is similar to that of chrysotile. Fig. 6 demonstrates the SEM photo showing the shape of asbestos fibers and the EDX composition chart. After determining that the fiber particles had the chemical composition of chrysotile by using EDX mounted on SEM, PCM was used to count the asbestos fibers among the air samples. 1/4 of the MCE filter paper was cut and preprocessed with the acetonetriacetin solution. The particle was magnified by 400 with the Walton-Backett graticule mounted on PCM, and asbestos fibers were counted within the graticule with the diameter of 100 mm on the microscope. Fibers of 5 mm or longer with a 3:1 ratio between length and width were counted, and sufficient number of graticule fields were used so that at least 100 fibers should be counted. The numbers of asbestos fiber particles and the concentration in the air sample were estimated with the following formula (NIOSH, 1994):
Ns ¼
C¼
NA an
NA a V n 1000
Where, Ns: The total number of fiber particles C: Concentration of asbestos fiber in the air (f/cc) N: The number of fiber particles counted A: The area from which the valid sample is collected (385 mm2) a: The area of graticule in which fiber particles are counted (0.0078 mm2) V: The amount of air sampled (L) n: The total number of graticule fields in which fiber particles are counted
94
Y.-C. Kim et al. / Journal of Cleaner Production 86 (2015) 88e97
Table 4 Results of the linear regression analysis of the amount of asbestos fiber generated per unit area of damaged slate.
Intercept Sheet Area
Coefficient
Standard error
0.077 0.159
0.008 9.683** 0.011 14.491** F ¼ 210.002** (P e value ¼ 0.000) R2(adjR2) ¼ 0.701 Regression equation of asbestos fiber concentration C ¼ 0.077 þ 0.159 A
t
Significance probability 0.000 0.000 * : p < 0.05,**: p < 0.01 C: Fiber Concentration A: Area of Slate
Fig. 7. Amount of asbestos fiber generated per unit area of damaged slate.
4. Estimation of the amount of asbestos fiber generated from damaged slates Sample results extracted by NIOSH 7400a are substituted to the formula mentioned in Section 3.2 to develop a database of asbestos fiber concentration of each air sample. It is then classified by evaluating the grade of the severity of damage mentioned in Section 2.2 to perform general analysis and regression analysis. Pearson correlation analysis was used to find out the area of asbestos slate that had an impact on the concentration of asbestos fibers. Finally the researchers conducted linear regression analysis with SPSS 20.0. 4.1. Analysis of asbestos fibers per unit area of slate In this research, 3 slates were selected for each of the 3 grades, and the air was sampled for each of the 9 slates; a total 90 air samples were analyzed with SEM/PCM. The average density of the slate was grade A (12.02 kg/m2) > grade B (11.60 kg/m2) > grade C (10.03 kg/m2). Area of the slate was between 0.089 m2 and 1.508 m2. This result shows that the grades of the slates assessed based on the damage as specified in paragraph 2.2 were appropriate. The PCM analysis showed that the density of asbestos fibers collected with the MCE filter was between 73.9 f/mm2 and 491.7 f/ mm2 and the concentration of asbestos fibers was between 0.051 f/ cc and 0.338 f/cc. The indoor air quality standard (0.01 f/cc) established by the Ministry of Environment was exceeded in all the 90 samples. Therefore, it was found that slates, damaged as they fall from the roof of a typical building in Korea, scatter asbestos fiber particles at a much higher concentration than the standard value, regardless of the severity of damage. Table 2 provides the results of the experiment.
Based on the analysis of 90 air samples, the researchers estimated the amount of asbestos fibers generated per unit area of the damaged slate. First, SPSS was used to find out that the area of asbestos slate had an impact on the concentration of asbestos fiber. According to Pearson correlation analysis, the Pearson correlation coefficient, which indicates the degree of correlation, was 0.839. This means a positive linear relationship was strong and a significant value at 0.01 in significance level. In short, the area of an asbestos slate is a factor that affects the concentration of asbestos fiber. A Pearson correlation analysis performed to each grade classified by the severity of the damaged slate found a correlation coefficient statistics to stand at 0.845, 0.901 and 0.900; which are all significant at 0.01 in significance level. Results of the Pearson correlation analysis are in Table 3 below. The researchers conducted linear regression analysis with SPSS 20.0, with the area of the slate as independent variable and the concentration of asbestos fiber as dependent variable. Results of the analysis are listed in Table 4. The analysis showed that the P value 0.000 was appropriate at the significance level of 0.05, and the modified R2 value was 0.701. When a slate is damaged as it falls freely from the roof height of a building, the concentration of asbestos fiber grows by 0.159 f/cc as the area of the slate grows by 1 m2. Fig. 7 illustrates the result of the analysis. 4.2. Analysis of the amount of asbestos fiber generated per unit area of slate by grade The researchers calculated the amount of asbestos fiber generated per unit area of damaged slate by the grade of damage as deduced in paragraph 2.2. Based on the results of the analysis of 30 air samples for each of the 3 grades (A, B, C), the researchers conducted a linear regression analysis with SPSS 20.0, with the area of
Y.-C. Kim et al. / Journal of Cleaner Production 86 (2015) 88e97
95
Table 5 Results of the linear regression analysis of the amount of asbestos fiber generated per unit area of damaged slate (Group A). Group A
Coefficient
Standard error
Intercept Sheet Area
0.080 0.127
0.010 8.040** 0.015 8.350** F ¼ 69.724** (P e value ¼ 0.000) R2(adjR2) ¼ 0.703 Regression equation of asbestos fiber concentration CA ¼ 0.080 þ 0.127 AslateA
t
Significance probability 0.000 0.000 * : p < 0.05,**: p < 0.01 CA: Fiber Concentration (A Group) AslateA: Area of Slate (A Group)
Table 6 Results of the linear regression analysis of the amount of asbestos fiber generated per unit area of damaged slate (Group B). Group B
Coefficient
Standard error
Intercept Sheet Area
0.064 0.157
0.012 5.283** 0.014 10.992** F ¼ 120.833** (P e value ¼ 0.000) R2(adjR2) ¼ 0.805 Regression equation of asbestos fiber concentration CB ¼ 0.064 þ 0.157 AslateB
t
Significance probability 0.000 0.000 * : p < 0.05,**: p < 0.01 CB: Fiber Concentration (B Group) AslateB: Area of Slate (B Group)
Table 7 Results of the linear regression analysis of the amount of asbestos fiber generated per unit area of damaged slate (Group C). Group C
Coefficient
Standard error
Intercept Sheet Area
0.069 0.221
0.013 5.310** 0.020 10.943** F ¼ 119.746** (P e value ¼ 0.000) R2(adjR2) ¼ 0.804 Regression equation of asbestos fiber concentration CC ¼ 0.069 þ 0.221 AslateC
the slate as independent variable and the concentration of asbestos fiber as dependent variable. For grade A slates, the analysis showed that the P value 0.000 was appropriate at the significance level of 0.05, and the modified R2 value was 0.703. This shows that for grade A slates, the concentration of asbestos fiber grows by 0.127 f/cc as the damaged area grows by 1 m2. For grade B slates the analysis showed that the P value 0.000 was appropriate at the significance level of 0.05, and the modified R2 value was 0.805. This shows that for grade B slates, the concentration of asbestos fiber grows by 0.157f/cc as the damaged area grows by 1 m2. For grade C slates the analysis showed that the P value 0.000 was appropriate at the significance level of 0.05, and the modified R2 value was 0.804. This shows that for grade C slates the concentration of asbestos fiber grows by 0.221 f/cc as the damaged area grows by 1 m2. Tables 5e7 show the result of regression analysis for each grade, and the regression expression about the amount of asbestos fiber generated per unit area by grade. 4.3. Comparison of the amount of asbestos fiber generated from the slate between grades The amount of asbestos fiber generated from the damaged slate was compared between grades A, B and C. Fig. 8 illustrates the amount of asbestos fiber generated per unit area of the damaged slate by grade of damage. The result of the analysis in paragraph 4.2 shows that the concentration of asbestos fiber per unit area was 0.127 f/cc for grade A, 0.157 f/cc for grade B, and 0.221 f/cc for grade C. This shows that more asbestos fiber is generated as the slate is more damaged. When the amount of asbestos fiber generated by grade is compared with the analysis of paragraph 4.1 (average amount of asbestos fiber per unit area; 0.159 f/cc), grade C slates generate more asbestos fiber than the average by about 39.0%, and grade A generates less by about 20.1%.
t
Significance probability 0.000 0.000 * : p < 0.05,**: p < 0.01 CC: Fiber Concentration (C Group) AslateC: Area of Slate (C Group)
5. Conclusions Use of asbestos is banned or restricted around the world but there still exists a large amount of asbestos slates. An effective maintenance, disintegration and removal of asbestos slates during normal times are critical. Of equal importance is quantifying the amount of asbestos generated during disaster situations like earthquakes as a way to develop solutions to keep asbestos from scattering. While many preceding research focused on the harmfulness of asbestos and asbestos release from slates, so far no research has looked into asbestos fiber generated from a damaged slate. This raises the need to quantify the amount of asbestos fiber generated when a slate is damaged. This research estimated the amount of asbestos fiber generated per unit area of damaged slates by severity of damage. Used slates were collected from buildings and divided into three grades depending on the severity of damage. To simulate natural damage, the researchers free-dropped slates from the roof height of a typical building and sampled the air surrounding it. The asbestos fiber particles were counted with SEM/PCM in the air samples and classified into three grades. Based on the results, the researchers conducted a regression analysis with SPSS 20.0, and developed a model for estimating the amount of asbestos fiber generated per unit area of damaged slate. SEM/PCM analysis of air samples showed that damaged slates emitted chrysotile, and the content of asbestos fiber emitted to the air was between 0.051 f/cc and 0.338 f/cc. It was found that the indoor air quality standard of 0.01 f/cc established by the Ministry of Environment was exceeded in all 90 samples. This shows that damaged slates are very dangerous as they emit a large amount of asbestos fibers. Linear regression analysis showed that damaged slates generate asbestos fibers of 0.159 f/cc per unit area. Tables 5e7 show the regression expressions for the amount of asbestos fiber generated by grade. The generation of asbestos fiber per unit area
96
Y.-C. Kim et al. / Journal of Cleaner Production 86 (2015) 88e97
Fig. 8. Amount of asbestos fiber generated per unit area of damaged slate by grade of damage.
was 0.127 f/cc for grade A, 0.157 f/cc for grade B, and 0.221 f/cc for grade C, which shows that more asbestos fiber is generated as damage of the slate worsens. The researchers developed the asbestos fiber calculation formula for each grade. When the amount of asbestos fiber generated by grade is compared with the average amount of asbestos fiber per unit area, grade C slate generates more asbestos fiber than the average by about 39.0%, and grade A generates less by about 20.1%. This research revealed that asbestos materials used in buildings are far more hazardous when damaged than when they are not. It was found that, the greater the damage, more asbestos fiber would scatter. The estimation model offered in this study can quantify the amount of asbestos fiber based on the building area by estimating the amount of asbestos fiber scattered due to the slate's severity of damage. This estimation model is expected to contribute to the quantification of asbestos generated due to disasters, including earthquakes, by building and by area, which were not been possible before due to the lack of relevant studies. A formula to determine the amount of asbestos fiber from this research is needed at a time when research to control asbestos is increasingly becoming a critical issue around the world. It is very important to quantify asbestos fiber emitted due to disasters and formulate measures to prevent scattering as to maintain, dismantle and remove asbestos slates effectively in normal times. In this sense, the asbestos fiber estimating formula suggested in this study is deemed important. The result of this study may be utilized as basic data to estimate the influence of the slate present in buildings on the environment and human body, and as a guideline to prevent damages to asbestos and to handle asbestos in case of disaster. In future research, a more indepth research evaluating ambient environment, which accounts for air movement, and risk to the human body based on findings of this research to develop a simulation on the spread of asbestos fiber will hopefully support the efficient use of the simulation in disaster situations.
Acknowledgments This research was supported by a grant (12 High-tech Urban C20) from High-tech Urban Development Program funded by Ministry of Land, Infrastructure and Transport of Korea government (12 High-tech Urban C20).
References AHERA (Asbestos Hazard Emergency Response Act), 2003, 1987. Part 763, Subpart E Asbestos-containing Materials in Schools. EPA, 52 CFR Part 763. http://www2. epa.gov/sites/production/files/documents/2003pt763_0.pdf. ASTM (American Society for Testing and Materials), 2014. Standard Practice for Comprehensive Building Asbestos Surveys. DOI.10.1520/E2356-10. http://www. astm.org/Standards/E2356.htm. Bae, Jae-Min, Kim, Young-Chan, Hong, Won-Hwa, 2013. The development of estimation model for calculated CO2 emissions from dismantling and demolition of asbestos containing materials. Archit. Inst. Korea 29 (10), 267e274. Becklake, M.R., 1976. Asbestos-related diseases of the lung and other organs: their epidemiology and implications for clinical practice. Am. Rev. Respir. Dis. 1, 187e227. Beddoe, R.E., Dorner, H.W., 2005. Modelling acid attack on concrete: part I. The essential mechanisms. Cem. Concr. Res. 12, 2333e2339. Bornemann, P., Hildebrandt, U., 1986. On the problem of environmental pollution by weathering products of asbestos cement. Staub Reinhalt. Luft 487e489. Choi, Jung-Keun, Paek, Do-Myung, Paik, Nam-Won, 1998. The production, the use, the number of workers and exposure level of asbestos in Korea. Korean Ind. Hyg. Assoc. J. 8 (2), 242e253. Deng, Q., Lan, Y., Wang, M., 2009. 30 years follow up study: the dose-response relationship of asbestos exposure and asbestosis. Mod. Prev. Med. 36 (11), 2027e2029. Doll, R., Peto, J., Britain, G., 1985. Effects on Health of Exposure to Asbestos. Health & Safety Commission. Dyczek, J., 2006. Surface of asbestos-cement (AC) roof sheets and assessment of the risk of asbestos release. In: Conference Presentation Krakow, Poland, September 28e29. Ernst, W., 2012. Overview of naturally occurring earth materials and human health concerns. J. Asian Earth Sci. 108e126. Favero-Longo, S., Castelli, D., Fubini, B., Piervittori, R., 2009. Lichens on asbestosecement roofs: bioweathering and biocovering effects. J. Hazard. Mater. 2, 1300e1308. Hourihane, D.O., 1964. The pathology of mesotheliomata and an analysis of their association with asbestos exposure. Thorax 3, 268e278. Health and Safety Executive (HSE), 1998. MDHS 87 Fibres in Air: Guidance on the Discrimination Between Fibre Types in Samples of Airborne Dust on Filters Using Microscopy. http://www.hse.gov.uk/pubns/mdhs/pdfs/mdhs87.pdf. HSE (Health and Safety Executive), 2002. A Comprehensive Guide to Managing Asbestos in Premises. ISBN 978 0 7176 2381 5/HSG227. http://www.hse.gov.uk/ pubns/priced/hsg227.pdf. Jinhui, L., Qingyin, D., Keli, Y., Lili, L., 15 October 2014. Asbestos and asbestos waste management in the Asian-Pacific region: trends, challenges and solutions. J. Clean. Prod. 81, 218e226. Kamp, D.W., 2009. Asbestos-induced lung diseases: an update. Transl. Res. 4, 143e152. Kane, A.B., Boffetta, P., Saracci, R., Wilbourn, J., 1996. Mechanisms of Fibre Carcinogenesis. International Agency for Research on Cancer. Kazan-Allen, L., 2005. Asbestos and mesothelioma: worldwide trends. Lung Cancer S3eS8. Kim, Ji-Yeong, Lee, Song-Kwon, Lee, Jeong-Hee, Lim, Mu-Heok, Kang, Sung-Wook, Phee, Young-Gyu, 2009. Study on the factors affecting asbestos exposure level from asbestos abatement in building. J. Korean Soc. Occup. Environ. Hyg. 19 (1), 8e15.
Y.-C. Kim et al. / Journal of Cleaner Production 86 (2015) 88e97 Kim, Hyun-Wook, Park, Gye-Young, Han, Jin-Gu, Han, Young-Sun, Hwang, Bum-Gu, Lee, Jun-Hyuk, 2010. Releasing of asbestos fibers from the weathered asbestos cement slate roofing. J. Korean Soc. Occup. Environ. Hyg. 20 (2), 88e93. Kim, Y., Son, B., Park, W., Hong, W., 2011a. A study on the distribution of the asbestos cement slates and calculation of disposal cost in the rural area. Archit. Res. 2, 31e40. Kim, Young-Chan, Son, Byeung-Hun, Kim, Hye-Mi, Hong, Won-Hwa, 2011b. A study on the distribution maps for asbestos cement slates using GIS. J. Korea Inst. Ecol. Archit. Environ. 11 (3), 57e62. Kim, Young-Chan, Hong, Won-Hwa, Zhang, Yuan-Long, 2014. A study on the efficient method of information offering through survey of asbestos awareness of deteriorated apartment facility managers. Archit. Inst. Korea 30 (1), 217e224. Korea Occupational Safety and Health Agency, 2006. KOSHA CODE A-1-2004. Lee, Seung-Ho, 2002. The characteristics of folk houses at heavy snowfall regions in Korea. J. Korean Geogr. Soc. 37 (1), 1e14. Linton, A., Vardy, J., Clarke, S., van Zandwijk, N., 2012. The ticking time-bomb of asbestos: its insidious role in the development of malignant mesothelioma. Crit. Rev. Oncol. 2, 200e212. Meyer, E., 1986. Examination of the importance of weathering asbestos cement planes for concentration of asbestos fibers in the environment. Staub Reinhalt. Luft 482e483. Ministry of Government Legislation, 2014. Asbestos Safety Management Act. http:// www.law.go.kr/%EB%B2%95%EB%A0%B9/%EC%84%9D%EB%A9%B4%EC%95%88% EC%A0%84%EA%B4%80%EB%A6%AC%EB%B2%95. Nicholson, W.J., 2001. The carcinogenicity of chrysotile asbestosea review. Ind. Health 2, 57e64. National Institute for Occupational Safety and Health (NIOSH), 1994. Asbestos and Other Fibers by PCM: 7400, fourth ed. In: NIOSH Manual of Analytical Methods
97
(NMAM) S. Department of HHS http://www.cdc.gov/niosh/docs/2003-154/pdfs/ 7400.pdf. Occupational Safety and Health Administration (OSHA), 1997. OSHA ID-160, Asbestos in Air. OSHA. https://www.osha.gov/dts/sltc/methods/inorganic/ id160/id160.pdf. Pastuszka, J.S., 2009. Emission of airborne fibers from mechanically impacted asbestos-cement sheets and concentration of fibrous aerosol in the home environment in Upper Silesia, Poland. J. Hazard. Mater. 162 (2), 1171e1177. Selikoff, I.J., Churg, J., Hammond, E.C., 1964. Asbestos exposure and neoplasia. JAMA 1, 142e146. Sim, M.R., 2013. A worldwide ban on asbestos production and use: some recent progress, but more still to be done. Occup. Environ. Med. 1, 1e2. Spurny, K., 1989. On the release of asbestos fibers from weathered and corroded asbestos cement products. Environ. Res. 1, 100e116. The Ministry of Environment of Korea government, 2009. Asbestos Management Catalogue. http://www.me.go.kr/. Turci, F., Favero-Longo, S.E., Tomatis, M., Martra, G., Castelli, D., Piervittori, R., Fubini, B., 2007. A biomimetic approach to the chemical inactivation of chrysotile fibres by lichen metabolites. Chem. A Eur. J. 14, 4081e4093. U.S. EPA (Environmental Protection Agency), 2003. Managing Asbestos in Place: a Building Owner's Guide to Operations and Maintenance Programs for Asbestoscontaining Materials (Green Book). Office of Pesticides and Toxic Substances, Washington, DC. http://nepis.epa.gov/Exe/ZyPDF.cgi/20011E4D.PDF? Dockey¼20011E4D.PDF. Virta, R.L., 2006. Worldwide Asbestos Supply and Consumption Trends from 1900 Through 2003. US Geological Survey Reston, VA. Wagner, J., 1965. The sequelae of exposure to asbestos dust. Ann. N. Y. Acad. Sci. 1, 691e695.