- Email: [email protected]
Identification and quantification of indoor air pollutant sources within a residential academic campus
Science of the Total Environment 569–570 (2016) 46–52
Contents lists available at ScienceDirect
Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv
Identification and quantification of indoor air pollutant sources within a residential academic campus Shalini Suryawanshi a, Amit Singh Chauhan a, Ritika Verma a, Tarun Gupta a,b,⁎ a b
Department of Civil Engineering, Indian Institute of Technology Kanpur, India APTL, Centre for Environmental Science and Engineering, Indian Institute of Technology Kanpur, India
H I G H L I G H T S
G R A P H I C A L
A B S T R A C T
• Average PM0.6 concentration indoor was about 94.44 μg/m3. • PMF used to identify the strength of indoor air pollution sources. • Eight metals Ba, Ca, Cr, Cu, Fe, Mg, Ni and Pb were quantified • Identified sources: Coal combustion, Tobacco smoking, Wall dust, Soil particles & Wooden furniture.
a r t i c l e
i n f o
Article history: Received 23 March 2016 Received in revised form 23 May 2016 Accepted 10 June 2016 Available online xxxx Editor: D. Barcelo Keywords: Source apportionment ICP-OES PMF Indoor air pollution Metals
⁎ Corresponding author. E-mail address: [email protected] (T. Gupta).
http://dx.doi.org/10.1016/j.scitotenv.2016.06.061 0048-9697/© 2016 Elsevier B.V. All rights reserved.
a b s t r a c t There is a growing concern regarding the adverse health effects due to indoor air pollution in developing countries including India. Hence, it becomes important to study the causes and sources of indoor air pollutants. This study presents the indoor concentrations of PM0.6 (particles with aerodynamic diameter less than 0.6 μm) and identifies sources leading to indoor air pollution. Indoor air samples were collected at IIT Kanpur campus. Ninety-eight PM0.6 samples were collected during November 2013 to September 2014. PM0.6 concentration was measured using a single stage impactor type PM0.6 sampler. The average PM0.6 concentration indoor was about 94.44 μg/m3. Samples collected were then analysed for metal concentrations using ICP-OES (Inductively Coupled Plasma – Optical Emission Spectrometer). Eight metals Ba, Ca, Cr, Cu, Fe, Mg, Ni and Pb were quantified from PM samples using ICP-OES. Positive Matrix Factorization (PMF) was used for source apportionment of indoor air pollution. PMF is a factor analysis tool which helps in resolving the profile and contribution of the sources from an unknown mixture. Five possible sources of indoor pollutants were identified by factor analysis – (1) Coal combustion (21.8%) (2) Tobacco smoking (9.8%) (3) Wall dust (25.7%) (4) Soil particles (17.5%) (5) Wooden furniture/paper products (25.2%). © 2016 Elsevier B.V. All rights reserved.
S. Suryawanshi et al. / Science of the Total Environment 569–570 (2016) 46–52
1. Introduction Indoor air pollution started back in the primitive times when people settled at one place in community and started using fire for warmth, cooking and light. In India about 32% of the primary energy needs are still derived from the use of biomass and about 70% of the India's population depends on biomass for energy needs (http://mnre.gov.in/ schemes/grid-connected/biomass-powercogen/). Biomass fuel burning is one of the major sources of indoor air pollution. Indoor air pollution refers to the poor quality of air within and around the structures and buildings. It leads to the health problems and discomfort for the occupants of the building. Most of the indoor air pollutants are produced within the building whereas some of the pollutants might enter from outside due to improper ventilation. Health effects due to poor indoor air quality generally occurs after years of exposure to pollutants but some health effects like irritation in throat, eyes and nose, dizziness and headache might occur even after a short term exposure to the pollutants. A debate is still going on whether the size, chemical composition or surface area is most relevant in predicting the health effects due to particulate matter (Lighty et al., 2000). Studies have shown that increased amount of particulate matter present in the air have a huge impact on the health of the exposed population (Dockery et al., 1993; Koenig et al., 2005). A survey showed that employed persons in U.S. spend only about 2% of their time outdoors, 6% of their time in transit, and 92% of their time indoors (Klepeis et al., 2001). Since people spend most of their time indoor so indoor air pollutants have a huge impact on the health of the people. Personal exposure to pollutants depends on the time spent indoor and concentration of particulate matter in air. Since women and children spend most of their time indoor, they are more exposed to indoor pollutants. According to WHO about 4.3 million people die every year due to household air pollutants (http://www.who.int/mediacentre/factsheets/fs292/en/). Indoor air pollutants concentrations are higher than outdoor. Approximately 76% of the particulate matter are present indoor in the developing countries (Smith, 1993). Children below five years who spend a lot of time indoor have high risk of lower respiratory disorders (Smith et al., 2000) like pneumonia (Dherani et al., 2008). More than 50% of the deaths of children below 5 years of age from acute lower respiratory infection are due to the household indoor air pollutants from solid fuel burning (http:// www.who.int/mediacentre/factsheets/fs292/en/). Lung and cardiovascular diseases are common among the people living in places having high concentration of indoor pollutants (Perez-Padilla et al., 2010; Uzoigwe
47
et al., 2013). About 17% of deaths among adults due to lung cancer are because of the carcinogens present in air due to household air pollution (http://www.who.int/mediacentre/factsheets/fs292/en/). India has the highest negative impact of indoor air pollution in the households. According to statistics, India alone is responsible for 28% deaths due to indoor air pollution in the developing countries (Rohra and Taneja, 2016). India ranks second in the world for per capita burden of the respiratory illness (Smith, 2000). This makes it necessary to study the sources of indoor pollutants and implement effective control measures. 96 PM0.6 samples were collected from different indoor microenvironments in this study. The collected samples were further subjected to chemical analysis. The metals analysed in this study are Fe, Pb, Ba, Ca, Cr, Cu, Mg and Ni. The objective of this study is source apportionment of the sources of indoor air pollutants. Source apportionment was carried out using Positive Matrix Factorization (Norris and Duvall, 2014). PMF is a factor analysis tool which resolves the contributions and profiles of the sources in an unknown mixture. PMF does not analyse on sample by sample basis like many other factor analysis tools and so has less chances of error in source contribution due to variations that might occur in the source profiles due to various reasons. PMF uses all the data together and gives the average source profile over the time period of collection of samples (Hopke, 2014). 2. Materials and methods 2.1. Site location The study was conducted in IIT Kanpur campus (26.5114° N, 80.2349° E, and 142 m above the mean sea level). IIT Kanpur is situated beside the Grand Trunk road. Since it is situated on the outskirts of Kanpur city, it has a cleaner environment as compared to the rest of the city (Devi et al., 2009). Indoor air sampling was done at three places, Hall 4 room, Atmospheric Particle Technology laboratory (APTL) and Western laboratories. 2.2. Sampling procedure Particulate matter was collected using a single stage impactor type PM0.6 sampler (standardized against Aerodynamic Particle Sizer, APS). Flow rate of sampler was 12 LPM which was controlled by the rotameter (calibrated using a mass flow meter). Polytetrafluoroethylene (PTFE) or Teflon filter papers with 47 mm diameter and 2 μm pore size were used
Fig. 1. Variation of PM0.6 and metal concentrations with time.
48
S. Suryawanshi et al. / Science of the Total Environment 569–570 (2016) 46–52
to collect PM0.6 particles (Chauhan and Gupta, 2016). The design and performance related details of the sampler are provided in the supplementary section. The reasons for using Teflon filter paper are: i. Teflon filter paper is chemically inert and has very little impurities and thus it is suitable for trace element analysis. ii. It has high collection efficiency and very little moisture absorption capacity. iii. It can be used in a wide range of weather conditions without any deformation.
Table 2 MDLj and pj values of elements for uncertainity calculation (Kim et al., 2005). Species
MDLj value (ng/m3)
pj value (%)
Ba Ca Cr Cu Fe Mg Ni Pb
0.11 0.11 4.13 10.42 6.25 0.1 2.08 22.928
10 22 10 10 10 10 10 10
2.3. Collection and storage of samples Since the PM mass collected is in very small quantity so utmost care was taken for filter preparation, sample collection and in weighing filter papers. Filter paper used in sampling was pre-conditioned in a controlled environment at 25 °C and 50% RH for about 24 h and same procedure was followed after the completion of sampling so that there is no error in the PM0.6 mass due to moisture. Each pre-conditioned filer was then weighed three times using Mettler balance (APM 440, Mettler, sensitivity 5 μg) and if the difference between any readings was found to be more than 5 μg then the filter paper was re-conditioned and reweighed (Chakraborty and Gupta, 2010). The filter papers were transferred to sealed plastic containers and then according to the sampling schedule the filter papers were placed on the sampler using clean forceps. The filter papers were taken out of the samplers using clean forceps after the sampling and were transferred back to the sealed plastic containers. Collected filter papers were post-conditioned under similar controlled environment as during pre-conditioning (25 °C, 50% RH) and weighed to calculate PM0.6 mass. After weighing, the filter papers were transferred to plastic containers and kept in refrigeration at 5 °C till further analysis. For every month about 10% of the filter papers were kept as blank and they were subjected to similar procedure as the filter papers which were used during sampling. 2.4. Calculation of PM0.6 mass Ninety-eight PM0.6 samples were collected during November 2013 to September 2014. PM0.6 mass was calculated by the difference in the mass of the filter papers after and before sampling. PM0.6 concentration was then calculated by dividing the PM0.6 mass by the volume of air sampled. PM0.6 concentration was calculated using the following equation: W f −W i −W b 106 PM0:6 concentration μg=m3 ¼ V
ð1Þ
2.5. Analysis of elements For the purpose of elemental analysis, Teflon filter paper was cut using plastic scissors into several small pieces and was placed at the bottom of round bottom glass digestion vessel. 20 ml of concentrated Nitric acid (65%, GR grade, Merck Suprapure) was poured into the digestion vessel and for 2 h the digestion vessel was put on hot plate having 180 °C temperature. It was taken care that the digestion vessel did not completely dry up because of the evaporation of Nitric acid and if required extra Nitric acid was added during digestion. After digestion the vessel was allowed to cool down to room temperature. The walls of the vessel were rinsed with Milli-Q water (resistivity = 18.2 MΩ cm at 25 °C) and kept for 30 min for diffusion of acid from the filters into the solution, then the solution in the vessel was filtered through 0.22 μm filters and the final volume was made up to 100 ml using Milli-Q water. These aqueous samples were then analysed for elements using ICP-OES (Inductively Coupled Plasma – Optical Emission Spectrometer, Thermo Fisher, iCAP 6300 Duo) as per the earlier established procedures (Chakraborty and Gupta, 2010). Blank filters were also subjected to the same procedure. 2.6. ICP-OES ICP-OES is generally used for low to moderate detection limit i.e., 0.1– 100 ppb. The detection limit of elements used in this analysis is shown in Table 2. In ICP-OES heated Argon plasma is used to excite the concerned element's atom in the sample and then spectrometer measures the intensity of the emitted electromagnetic radiations during transformation from higher to lower energy state at a characteristic wavelength for each element. The measured intensity is converted into concentration using the calibration curve of that metal generated using NIST traceable external multi-element standard solution (Boss and Fredeen, 2004; Gupta and Mandariya, 2013). Ba, Ca, Cr, Cu, Fe, Mg, Ni and Pb concentrations were quantified in the indoor air samples with ICP-OES.
Where, 2.7. Positive matrix factorization Wf = Mass of filter paper after sampling (grams). Wi = Mass of filter paper before sampling (grams). Wb = Mass increase of blank filter (grams). V = Volume of sampled air (m3)
US EPA PMF version 5.0.14 software was used for source apportionment. PMF requires a data set consisting of certain parameters for different samples. It also requires uncertainty data for individual parameters
Table 1 Statistical analysis of the metal concentrations (μg/m3) measured during the study period.
Table 3 PMF analysis results for PM0.6 (μg/m3).
Species
Category
Signal/noise
Min
25th
50th
75th
Max
Species
Factor 1
Factor 2
Factor 3
Factor 4
Factor 5
Ba Ca Cr Cu Fe Mg Ni Pb
Strong Strong Strong Strong Strong Strong Strong Weak
6.6 6.5 4.8 5.0 5.0 6.3 6.5 3.6
0.009 0.385 0.015 0.009 0.055 0.054 0.004 0.018
0.241 3.930 0.522 0.675 2.707 1.834 0.219 0.699
0.619 6.021 0.745 0.989 2.823 2.685 0.410 0.699
1.176 7.955 0.877 1.284 3.576 3.303 0.699 0.699
6.605 16.800 5.574 5.320 9.869 12.558 4.500 18.408
Ba Ca Cr Cu Fe Mg Ni Pb
1.137 1.095 0.000 0.000 1.453 0.005 0.000 0.000
0.000 0.693 0.000 0.043 0.000 0.301 0.613 0.000
0.000 2.036 0.000 1.580 0.481 0.080 0.022 0.137
0.175 0.000 0.230 0.178 0.000 2.304 0.076 0.000
0.000 1.481 0.963 0.000 1.087 0.353 0.000 0.373
S. Suryawanshi et al. / Science of the Total Environment 569–570 (2016) 46–52
49
Fig. 2. Percentage contribution of elements to different factors (Factor fingerprints).
in samples. Uncertainty is calculated using the following formula (Kim et al., 2005) uij ¼ p j xij þ MDL j Where,
3
ð2Þ
uij = uncertainty. pj = parameter proportional of uncertainty. xij = concentration of specie j in sample i. MDLj = Minimum Detection Limit. Positive Matrix Factorization is based on multivariate factor analysis which breaks down a matrix having species data into two matrices; factor contribution and factor profile. The factor profile
Fig. 3. Factor profiles.
50
S. Suryawanshi et al. / Science of the Total Environment 569–570 (2016) 46–52
Fig. 4. Source apportionment of indoor air pollutants.
needs interpretation by the user to identify the sources of pollutants which are contributing to the measured data by comparing the factor profile data with emissions and discharges from inventories/other sources. p
xij ¼ ∑ g ik f kj þ eij k¼1
ð3Þ
j - specie. Factor profile and contributions are obtained in PMF by minimising the objective function Q. " #2 p xij − ∑k¼1 g ij f kj Q ¼ ∑∑ uij i¼1 j¼1 n
m
ð4Þ
Where, xij - concentration of a species. eij - residual matrix. fkj - factor profile. gik - factor contributions. p - factor number. i - sample number.
Where, uij is uncertainty of specie j in sample i. There are two types of Q displayed by the model results, Q robust and Q true. Q true is calculated using all the points whereas Q robust is calculated by excluding the points which are not fit by the model, having uncertainty scaled residual greater than 4. The difference between Q robust and Q true shows the effect of inconsistencies from sources which are present during the sampling period (Norris and Duvall, 2014).
Fig. 5. Factor contributions (μg/m3) to the PM0.6 mass.
S. Suryawanshi et al. / Science of the Total Environment 569–570 (2016) 46–52
For choosing the number of factors, Q = i ∗ j is a good starting point. The model interpretation, Q values, the goodness of model fit and post PMF regression are all considered together to find the optimum number of factors. PMF require multiple iterations, it starts the search with a randomly generated factor profile. Since the starting point is random in nature there is no guarantee that PMF will give a global minimum, it may instead find a local minimum. So in order to find a global minimum PMF should be run multiple times iteratively. There can exist a multiple number of F (factor profile) and G (factor contribution) matrices giving the same minimum value of Q. This is known as rotational ambiguity. In order to check rotational ambiguity PMF has FPEAK tool. Five FPEAK runs could be performed in PMF for the current dataset. PMF accepts only non-negative values. It does not work with the missing data in data matrix. In case, if missing values are present then there are three options – remove the entire sample, remove the species or replace missing value with mean value. Measurements which are not relevant to the sources in study are discarded from the PMF analysis. Data below detection limit (DL) is replaced by DL/2 and its uncertainty is replaced by (5/6) DL. Species can also be categorised as bad, weak and strong based on signal to noise ratio and their uncertainties are also changed accordingly (Hopke, 2014). 3. Results and discussions PM0.6 and metal concentration varies a lot on different days as shown in Fig. 1. The maximum and minimum PM0.6 concentration was found out to be 330.79 μg/m3 and 17.59 μg/m3, respectively. The average PM0.6 concentration was 94.44 μg/m3. The statistical analysis of metal concentrations measured during the study period (monthly average) is shown in Table 1. PMF analysis was carried out on elemental concentrations taken over a period of time. The uncertainty values were calculated using Eq. (2) and values as depicted in Table 2. Several parameters were considered for obtaining the result. The missing values in the data were replaced by species median. Certain species were considered as bad species in PMF analysis and their uncertainty was replaced by a higher uncertainty value which was four times of the species-specific median. PMF was run for different number of factors. Q value decreased as the number of factors increased. Q theoretical for the data was approximately 784. Q (robust) for four and five factors was 1158.4 and 482.7, respectively. Pb was categorised as weak specie after analysing the base model run. F-peak analysis was also done to consider the rotational ambiguity. Five sources of indoor air pollutants were identified after PMF analysis (Table 3). G-Space plots in PMF showed that all the sources were independent and there was no correlation between them. Distribution of metal species in different sources shown in factor fingerprints plot (Fig. 2) gives the percentage contribution of metal species in different sources. It helps in the identification of major pollutant source releasing these metals. Factor 1 (Coal combustion) as shown in Fig. 3 had high concentration of Fe, Ba and Ca. A little amount of Mg was also present. These species are good indicator of coal combustion from households and from the power plants, brick kilns and industries present in and around the Kanpur city. According to a study Fe is present in the fly ash in coal in the form of Fe2O3 or Fe3O4 in about 2 to 26.8% (Cincinnati, 1969; Kang et al., 2011). Studies also shows release of CaO, Mg(OH)2 (Benson, 2006) and Ba (Kim and Kazonich, 1999; Kang et al., 2011) in coal combustion. Ba occurs in coal in different amount depending upon the type of coal (Zhao et al., 2014). Iron in coal is primarily associated with sulphur in minerals jarosite and pyrite, while other iron containing minerals like illite, ankerite and kaolinite also occurs in coal (Waanders et al., 2003). Calcium is generally present in coal as calcite (CaCO3) (Aladdin et al., 2013). A study showed that the ash deposits from coal fired power plant consists of CaO, MgO, Fe2O3 and other chemical compounds (Fernandez-Turiel et al., 2004). So this source indicates coal
51
combustion. The pollutants from industries are carried to the experimental site along with wind and enter the indoor environment through doors and windows. Factor 2 (Tobacco smoking) as shown in Fig. 3 had high concentration of Ca, Ni and significant amount of Mg, Cu. These pollutants are good indicators of tobacco smoking within indoor microenvironments. A study shows the effect of some important metals like Ca, Ni, Cu, Mg which are related to tobacco smoking (Chiba and Masironi, 1992). A study in China showed that both Chinese and imported cigarette papers had a significant amount of calcium and magnesium (Peng et al., 2011). Ni and Cu are toxic metals present in tobacco leaves (Bernhard et al., 2005). So this source indicates tobacco smoking. Factor 3 (Wall dust) as shown in Fig. 3 had high concentration of Ca, Cu and significant amount of Fe, Pb, Mg and Ni. Ca in the form of calcium carbonate is the main ingredient of wall putty used for finishing before painting the walls. According to studies, wall paints and construction materials are made up of metals like Pb, Zn, Cd, Cu, Mn, Ni, Co, Cr, Fe, etc. (Abagale and Campus, 2013; Mielke et al., 2001). So wall dust should be the source for this factor. Factor 4 (Soil particles) as shown in Fig. 3 had high concentration of Mg and significant amount of Cr, Cu, Ba and Ni. An USEPA study of 1983 shows the concentration of metals in the natural soil. Cr, Cu, Ba and Ni are present in the soil dust in approximately the same ratio as they are present in this factor (Lindsay, 1979). According to a study, soil consists of high concentration of Mg and a significant amount of Ba, Cu, Cr and Ni (Zeiner et al., 2015). So the source for factor 3 should be soil particles coming indoor along with wind or physical contact with people. Factor 5 (Wooden furniture/paper products) as shown in Fig. 3 had high concentration of Ca, Cr, Fe and significant amount of Mg and Pb. According to studies, wood has heavy metals like Ca, Mg, Al, Cd, Cr, Cu, Fe, Mn, Pb which may vary according to the climate and forest type (Arthur et al., 1999; Meier, 2013; Nicewicz and Szczepkowski, 2008). So the source of factor 5 could be the wooden furniture/paper products present indoor. The percentage share of different sources in PM0.6 (Fig. 4) shows that the maximum indoor pollution was caused due to wall dust (factor 3). Wall dust causes about 25.7% of indoor pollution. Wooden furniture/ paper products (factor 5) was the second highest pollutant source with a share of about 25.2%. Minimum indoor pollution was caused due to tobacco smoking (factor 2) with share of about 9.8%. Coal combustion (factor 1) and soil particles (factor 4) contributed about 21.8% and 17.5% respectively to indoor air pollution. Factor contributions of the sources were not constant and they changed with time (Fig. 5). This change in contribution of factors with time might be due to change in temperature, humidity and other influences.
4. Conclusions Sampling was done to collect PM0.6 concentrations at indoor sites in IIT Kanpur campus. Eight elements namely Ba, Ca, Cr, Cu, Fe, Mg, Ni and Pb were quantified using ICP-OES. A total of 98 samples were collected from November 2013 to September 2014 at suitable intervals. During this study it was found that there were large variations in PM0.6 concentrations with time PM0.6 concentration varies from minimum 17.59 μg/m3 to maximum 330.79 μg/m3. The average indoor PM0.6 concentration was 94.44 μg/m3. Source apportionment for the indoor air pollutants was done using PMF. PMF results gave source profile and source contributions of indoor air pollution. The analysis shows that five sources were responsible for the indoor pollution. The sources were independent of each other and there was no correlation between them. The source contribution showed variations with time. The five sources identified for indoor pollution were coal combustion (21.8%), tobacco smoking (9.8%), wall dust (25.7%), soil particles (17.5%) and wooden furniture/paper products (25.2%).
52
S. Suryawanshi et al. / Science of the Total Environment 569–570 (2016) 46–52
Acknowledgement This study was supported by the research fund from IIT Kanpur. References Abagale, S.A., Campus, N., 2013. Chemical Studies on the Composition of Natural Paint Pigment Materials from the Kassena-Nankana District of the Upper East region of Ghana. 3 pp. 13–23. Aladdin, M., Jian, J., Yang, Q., Chen, L.-C., Finkelman, R.B., Huang, X., 2013. Laboratory studies of the impact of calcite on in vitro and in vivo effects of coal dust: a potential preventive agent for coal workers' pneumoconiosis? Am. J. Ind. Med. 56, 292–299. Arthur, M.A., Siccama, T.G., Yanai, R.D., 1999. Calcium and magnesium in wood of northern hardwood forest species: relations to site characteristics. Can. J. For. Res. 29, 339–346. Benson, L.B., 2006. Use of Magnesium Hydroxide for Reduction of Plume Visibility in CoalFired Power Plants. Measurement. Bernhard, D., Rossmann, A., Wick, G., 2005. Metals in cigarette smoke. IUBMB Life 57, 805–809. Boss, C., Fredeen, K., 2004. Concepts, Instrumentation and Techniques in Inductively Coupled Plasma Optical Emission Spectrometry. PerkinElmer. Chakraborty, A., Gupta, T., 2010. Chemical Characterization and Source Apportionment of Submicron (PM1) Aerosol in Kanpur Region. Aerosol Air Qual. Res. 10, 433–445. Chauhan, Gupta, 2016. Development of a novel PM0.6 sampler (under review). Chiba, M., Masironi, R., 1992. Reviews/Analyses Toxic and trace elements in tobacco and tobacco smoke *. 70 pp. 269–275. Cincinnati, O., 1969. NEPIS document display. (URL:) http://nepis.epa.gov/Exe/ZyNET. exe/9100808I.TXT?ZyActionD=ZyDocument&Client=EPA&Index= Prior+to+1976&Docs=&Query=&Time=&EndTime=&SearchMethod= 1&TocRestrict=n&Toc=&TocEntry=&QField=&QFieldYear=&QFieldMonth= &QFieldDay=&IntQFieldOp=0&ExtQFieldOp=0&XmlQuery=&File=D% 3A\zyfiles\Index%20Data\70thru75\Txt\00000008\9100808I.txt&User= ANONYMOUS&Password=anonymous&SortMethod=h|-&MaximumDocuments= 1&FuzzyDegree=0&ImageQuality=r75g8/r75g8/x150y150g16/i425&Display= p|f&DefSeekPage=x&SearchBack=ZyActionL&Back=ZyActionS&BackDesc= Results%20page&MaximumPages=1&ZyEntry=1&SeekPage=x&ZyPURL (accessed 1.29.16). Devi, J.J., Gupta, T., Tripathi, S.N., Ujinwal, K., 2009. Assessment of Personal Exposure to Indoor and Outdoor Particulate Matter for Residents of an Academic Campus (IIT-Kanpur). Inhal. Toxicol. 21 (14), 1208–1222. Dherani, M., Pope, D., Mascarenhas, M., Smith, K.R., Weber, M., Bruce, N., 2008. Indoor air pollution from unprocessed solid fuel use and pneumonia risk in children aged under five years: a systematic review and meta-analysis. Bull. World Health Organ. 86, 390–394. Dockery, D.W., Pope 3rd, C.A., Xu, X., Spengler, J.D., Ware, J.H., Fay, M.E., Ferris, B.G.J., Speizer, F.E., 1993. An association between air pollution and mortality in six U.S. cities. N. Engl. J. Med. 329, 1753–1759. Fernandez-Turiel, J.L., Georgakopoulos, A., Gimeno, D., Papastergios, G., Kolovos, N., 2004. Ash deposition in a pulverized coal-fired power plant after high-calcium lignite combustion. Energy Fuel 18, 1512–1518. Gupta, T., Mandariya, A., 2013. Sources of Submicron Aerosol during Fog Dominated Wintertime at Kanpur. Environ. Sci. Pollut. Res. 20 (8), 5615–5629. Hopke, P.K., 2014. A Guide to Positive Matrix Factorization. pp. 1–16.
Kang, C.M., Gupta, T., Ruiz, P.A., Wolfson, J.M., Ferguson, S.T., Lawrence, J.E., Rohr, A.C., Goldleski, J., Koutrakis, P., 2011. Aged particles derived from emissions of coal-fired power plants: The TERESA field results. Inhal. Toxicol. 23, 11–30. Kim, A.G., Kazonich, G., 1999. Mass Release of Trace Elements from Coal Combustion ByProducts. Kim, E., Hopke, P.K., Qin, Y., 2005. Estimation of organic carbon blank values and error structures of the speciation trends network data for source apportionment. J. Air Waste Manage. Assoc. 55, 1190–1199. Klepeis, N.E., Nelson, W.C., Ott, W.R., Robinson, J.P., Tsang, a.M., Switzer, P., Behar, J.V., Hern, S.C., Engelmann, W.H., 2001. The National Human Activity Pattern Survey (NHAPS): a resource for assessing exposure to environmental pollutants. J. Expo. Anal. Environ. Epidemiol. 11, 231–252. Koenig, J.Q., Mar, T.F., Allen, R.W., Jansen, K., Lumley, T., Sullivan, J.H., Trenga, C.A., Larson, T.V., Liu, L.J.S., 2005. Pulmonary effects of indoor- and outdoor-generated particles in children with asthma. Environ. Health Perspect. 113, 499–503. Lighty, J.S., Veranth, J.M., Sarofim, A.F., 2000. Combustion aerosols: factors governing their size and composition and implications to human health. J. Air Waste Manage. Assoc. 50, 1522–1565. Lindsay, W.L., 1979. Chemical Equilibria in Soils. Wiley-Interscience Publication, Wiley. Meier, S., 2013. Elemental analysis of wood fuels NYSERDA's promise to New Yorkers : NYSERDA provides resources, expertise and objective information so New Yorkers can make confident, informed energy decisions. NYSERDA Rep. 13–13. Mielke, H.W., Powell, E.T., Shah, A., Gonzales, C.R., Mielke, P.W., 2001. Multiple metal contamination from house paints: consequences of power sanding and paint scraping in New Orleans. Environ. Health Perspect. 109, 973–978. Nicewicz, D., Szczepkowski, A., 2008. The content of heavy metals in the wood of healthy and dying beech trees (Fagus sylvatica L.). Acta Sci. Pol. Silv. Colendar. Rat. Ind. Lignar. 7 (4), 35–44. Norris, G., Duvall, R., 2014. EPA Positive M atrix Factorization (PM F) 5.0 Fundamentals and User Guide. Peng, W., Jun, S., Xiangmin, L., Lijun, Z., Li, L., 2011. Determination of calcium and magnesium in cigarette paper with autoanalyzer. Tob. Sci. Technol. 52–55. Perez-Padilla, R., Schilmann, A., Riojas-Rodriguez, H., 2010. Respiratory health effects of indoor air pollution. Int. J. Tuberc. Lung Dis. 14, 1079–1086. Rohra, H., Taneja, A., 2016. Indoor air quality scenario in India—an outline of household fuel combustion. Atmos. Environ. 129, 243–255. Smith, K., 1993. Fuel combustion, air pollution exposure, and health: the situation in developing countries. Annu. Rev. Energy Environ. 18, 529–566. Smith, K.R., 2000. National Burden of Disease in India From Indoor air Pollution. Smith, K.R., Samet, J.M., Romieu, I., Bruce, N., 2000. Indoor air pollution in developing countries and acute lower respiratory infections in children. Thorax 55, 518–532. Uzoigwe, J.C., Prum, T., Bresnahan, E., Garelnabi, M., 2013. The emerging role of outdoor and indoor air pollution in cardiovascular disease. N. Am. J. Med. Sci. 5, 445–453. Waanders, F.B., Vinken, E., Mans, A., Mulaba-Bafubiandi, A.F., 2003. Iron minerals in coal, weathered coal and coal ash - SEM and Mossbauer results. Hyperfine Interact. 148149, 21–29. Zeiner, M., Juranovic Cindric, I., Pozgaj, M., Pirkl, R., Silic, T., Stingeder, G., 2015. Influence of soil composition on the major, minor and trace metal content of Velebit biomedical plants. J. Pharm. Biomed. Anal. 106, 153–158. Zhao, C.L., Sun, Y.Z., Xiao, L., Qin, S.J., Wang, J.X., Duan, D.J., 2014. The occurrence of barium in a Jurassic coal in the Huangling 2 mine, Ordos Basin, northern China. Fuel 128, 428–432.
Contents lists available at ScienceDirect
Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv
Identification and quantification of indoor air pollutant sources within a residential academic campus Shalini Suryawanshi a, Amit Singh Chauhan a, Ritika Verma a, Tarun Gupta a,b,⁎ a b
Department of Civil Engineering, Indian Institute of Technology Kanpur, India APTL, Centre for Environmental Science and Engineering, Indian Institute of Technology Kanpur, India
H I G H L I G H T S
G R A P H I C A L
A B S T R A C T
• Average PM0.6 concentration indoor was about 94.44 μg/m3. • PMF used to identify the strength of indoor air pollution sources. • Eight metals Ba, Ca, Cr, Cu, Fe, Mg, Ni and Pb were quantified • Identified sources: Coal combustion, Tobacco smoking, Wall dust, Soil particles & Wooden furniture.
a r t i c l e
i n f o
Article history: Received 23 March 2016 Received in revised form 23 May 2016 Accepted 10 June 2016 Available online xxxx Editor: D. Barcelo Keywords: Source apportionment ICP-OES PMF Indoor air pollution Metals
⁎ Corresponding author. E-mail address: [email protected] (T. Gupta).
http://dx.doi.org/10.1016/j.scitotenv.2016.06.061 0048-9697/© 2016 Elsevier B.V. All rights reserved.
a b s t r a c t There is a growing concern regarding the adverse health effects due to indoor air pollution in developing countries including India. Hence, it becomes important to study the causes and sources of indoor air pollutants. This study presents the indoor concentrations of PM0.6 (particles with aerodynamic diameter less than 0.6 μm) and identifies sources leading to indoor air pollution. Indoor air samples were collected at IIT Kanpur campus. Ninety-eight PM0.6 samples were collected during November 2013 to September 2014. PM0.6 concentration was measured using a single stage impactor type PM0.6 sampler. The average PM0.6 concentration indoor was about 94.44 μg/m3. Samples collected were then analysed for metal concentrations using ICP-OES (Inductively Coupled Plasma – Optical Emission Spectrometer). Eight metals Ba, Ca, Cr, Cu, Fe, Mg, Ni and Pb were quantified from PM samples using ICP-OES. Positive Matrix Factorization (PMF) was used for source apportionment of indoor air pollution. PMF is a factor analysis tool which helps in resolving the profile and contribution of the sources from an unknown mixture. Five possible sources of indoor pollutants were identified by factor analysis – (1) Coal combustion (21.8%) (2) Tobacco smoking (9.8%) (3) Wall dust (25.7%) (4) Soil particles (17.5%) (5) Wooden furniture/paper products (25.2%). © 2016 Elsevier B.V. All rights reserved.
S. Suryawanshi et al. / Science of the Total Environment 569–570 (2016) 46–52
1. Introduction Indoor air pollution started back in the primitive times when people settled at one place in community and started using fire for warmth, cooking and light. In India about 32% of the primary energy needs are still derived from the use of biomass and about 70% of the India's population depends on biomass for energy needs (http://mnre.gov.in/ schemes/grid-connected/biomass-powercogen/). Biomass fuel burning is one of the major sources of indoor air pollution. Indoor air pollution refers to the poor quality of air within and around the structures and buildings. It leads to the health problems and discomfort for the occupants of the building. Most of the indoor air pollutants are produced within the building whereas some of the pollutants might enter from outside due to improper ventilation. Health effects due to poor indoor air quality generally occurs after years of exposure to pollutants but some health effects like irritation in throat, eyes and nose, dizziness and headache might occur even after a short term exposure to the pollutants. A debate is still going on whether the size, chemical composition or surface area is most relevant in predicting the health effects due to particulate matter (Lighty et al., 2000). Studies have shown that increased amount of particulate matter present in the air have a huge impact on the health of the exposed population (Dockery et al., 1993; Koenig et al., 2005). A survey showed that employed persons in U.S. spend only about 2% of their time outdoors, 6% of their time in transit, and 92% of their time indoors (Klepeis et al., 2001). Since people spend most of their time indoor so indoor air pollutants have a huge impact on the health of the people. Personal exposure to pollutants depends on the time spent indoor and concentration of particulate matter in air. Since women and children spend most of their time indoor, they are more exposed to indoor pollutants. According to WHO about 4.3 million people die every year due to household air pollutants (http://www.who.int/mediacentre/factsheets/fs292/en/). Indoor air pollutants concentrations are higher than outdoor. Approximately 76% of the particulate matter are present indoor in the developing countries (Smith, 1993). Children below five years who spend a lot of time indoor have high risk of lower respiratory disorders (Smith et al., 2000) like pneumonia (Dherani et al., 2008). More than 50% of the deaths of children below 5 years of age from acute lower respiratory infection are due to the household indoor air pollutants from solid fuel burning (http:// www.who.int/mediacentre/factsheets/fs292/en/). Lung and cardiovascular diseases are common among the people living in places having high concentration of indoor pollutants (Perez-Padilla et al., 2010; Uzoigwe
47
et al., 2013). About 17% of deaths among adults due to lung cancer are because of the carcinogens present in air due to household air pollution (http://www.who.int/mediacentre/factsheets/fs292/en/). India has the highest negative impact of indoor air pollution in the households. According to statistics, India alone is responsible for 28% deaths due to indoor air pollution in the developing countries (Rohra and Taneja, 2016). India ranks second in the world for per capita burden of the respiratory illness (Smith, 2000). This makes it necessary to study the sources of indoor pollutants and implement effective control measures. 96 PM0.6 samples were collected from different indoor microenvironments in this study. The collected samples were further subjected to chemical analysis. The metals analysed in this study are Fe, Pb, Ba, Ca, Cr, Cu, Mg and Ni. The objective of this study is source apportionment of the sources of indoor air pollutants. Source apportionment was carried out using Positive Matrix Factorization (Norris and Duvall, 2014). PMF is a factor analysis tool which resolves the contributions and profiles of the sources in an unknown mixture. PMF does not analyse on sample by sample basis like many other factor analysis tools and so has less chances of error in source contribution due to variations that might occur in the source profiles due to various reasons. PMF uses all the data together and gives the average source profile over the time period of collection of samples (Hopke, 2014). 2. Materials and methods 2.1. Site location The study was conducted in IIT Kanpur campus (26.5114° N, 80.2349° E, and 142 m above the mean sea level). IIT Kanpur is situated beside the Grand Trunk road. Since it is situated on the outskirts of Kanpur city, it has a cleaner environment as compared to the rest of the city (Devi et al., 2009). Indoor air sampling was done at three places, Hall 4 room, Atmospheric Particle Technology laboratory (APTL) and Western laboratories. 2.2. Sampling procedure Particulate matter was collected using a single stage impactor type PM0.6 sampler (standardized against Aerodynamic Particle Sizer, APS). Flow rate of sampler was 12 LPM which was controlled by the rotameter (calibrated using a mass flow meter). Polytetrafluoroethylene (PTFE) or Teflon filter papers with 47 mm diameter and 2 μm pore size were used
Fig. 1. Variation of PM0.6 and metal concentrations with time.
48
S. Suryawanshi et al. / Science of the Total Environment 569–570 (2016) 46–52
to collect PM0.6 particles (Chauhan and Gupta, 2016). The design and performance related details of the sampler are provided in the supplementary section. The reasons for using Teflon filter paper are: i. Teflon filter paper is chemically inert and has very little impurities and thus it is suitable for trace element analysis. ii. It has high collection efficiency and very little moisture absorption capacity. iii. It can be used in a wide range of weather conditions without any deformation.
Table 2 MDLj and pj values of elements for uncertainity calculation (Kim et al., 2005). Species
MDLj value (ng/m3)
pj value (%)
Ba Ca Cr Cu Fe Mg Ni Pb
0.11 0.11 4.13 10.42 6.25 0.1 2.08 22.928
10 22 10 10 10 10 10 10
2.3. Collection and storage of samples Since the PM mass collected is in very small quantity so utmost care was taken for filter preparation, sample collection and in weighing filter papers. Filter paper used in sampling was pre-conditioned in a controlled environment at 25 °C and 50% RH for about 24 h and same procedure was followed after the completion of sampling so that there is no error in the PM0.6 mass due to moisture. Each pre-conditioned filer was then weighed three times using Mettler balance (APM 440, Mettler, sensitivity 5 μg) and if the difference between any readings was found to be more than 5 μg then the filter paper was re-conditioned and reweighed (Chakraborty and Gupta, 2010). The filter papers were transferred to sealed plastic containers and then according to the sampling schedule the filter papers were placed on the sampler using clean forceps. The filter papers were taken out of the samplers using clean forceps after the sampling and were transferred back to the sealed plastic containers. Collected filter papers were post-conditioned under similar controlled environment as during pre-conditioning (25 °C, 50% RH) and weighed to calculate PM0.6 mass. After weighing, the filter papers were transferred to plastic containers and kept in refrigeration at 5 °C till further analysis. For every month about 10% of the filter papers were kept as blank and they were subjected to similar procedure as the filter papers which were used during sampling. 2.4. Calculation of PM0.6 mass Ninety-eight PM0.6 samples were collected during November 2013 to September 2014. PM0.6 mass was calculated by the difference in the mass of the filter papers after and before sampling. PM0.6 concentration was then calculated by dividing the PM0.6 mass by the volume of air sampled. PM0.6 concentration was calculated using the following equation: W f −W i −W b 106 PM0:6 concentration μg=m3 ¼ V
ð1Þ
2.5. Analysis of elements For the purpose of elemental analysis, Teflon filter paper was cut using plastic scissors into several small pieces and was placed at the bottom of round bottom glass digestion vessel. 20 ml of concentrated Nitric acid (65%, GR grade, Merck Suprapure) was poured into the digestion vessel and for 2 h the digestion vessel was put on hot plate having 180 °C temperature. It was taken care that the digestion vessel did not completely dry up because of the evaporation of Nitric acid and if required extra Nitric acid was added during digestion. After digestion the vessel was allowed to cool down to room temperature. The walls of the vessel were rinsed with Milli-Q water (resistivity = 18.2 MΩ cm at 25 °C) and kept for 30 min for diffusion of acid from the filters into the solution, then the solution in the vessel was filtered through 0.22 μm filters and the final volume was made up to 100 ml using Milli-Q water. These aqueous samples were then analysed for elements using ICP-OES (Inductively Coupled Plasma – Optical Emission Spectrometer, Thermo Fisher, iCAP 6300 Duo) as per the earlier established procedures (Chakraborty and Gupta, 2010). Blank filters were also subjected to the same procedure. 2.6. ICP-OES ICP-OES is generally used for low to moderate detection limit i.e., 0.1– 100 ppb. The detection limit of elements used in this analysis is shown in Table 2. In ICP-OES heated Argon plasma is used to excite the concerned element's atom in the sample and then spectrometer measures the intensity of the emitted electromagnetic radiations during transformation from higher to lower energy state at a characteristic wavelength for each element. The measured intensity is converted into concentration using the calibration curve of that metal generated using NIST traceable external multi-element standard solution (Boss and Fredeen, 2004; Gupta and Mandariya, 2013). Ba, Ca, Cr, Cu, Fe, Mg, Ni and Pb concentrations were quantified in the indoor air samples with ICP-OES.
Where, 2.7. Positive matrix factorization Wf = Mass of filter paper after sampling (grams). Wi = Mass of filter paper before sampling (grams). Wb = Mass increase of blank filter (grams). V = Volume of sampled air (m3)
US EPA PMF version 5.0.14 software was used for source apportionment. PMF requires a data set consisting of certain parameters for different samples. It also requires uncertainty data for individual parameters
Table 1 Statistical analysis of the metal concentrations (μg/m3) measured during the study period.
Table 3 PMF analysis results for PM0.6 (μg/m3).
Species
Category
Signal/noise
Min
25th
50th
75th
Max
Species
Factor 1
Factor 2
Factor 3
Factor 4
Factor 5
Ba Ca Cr Cu Fe Mg Ni Pb
Strong Strong Strong Strong Strong Strong Strong Weak
6.6 6.5 4.8 5.0 5.0 6.3 6.5 3.6
0.009 0.385 0.015 0.009 0.055 0.054 0.004 0.018
0.241 3.930 0.522 0.675 2.707 1.834 0.219 0.699
0.619 6.021 0.745 0.989 2.823 2.685 0.410 0.699
1.176 7.955 0.877 1.284 3.576 3.303 0.699 0.699
6.605 16.800 5.574 5.320 9.869 12.558 4.500 18.408
Ba Ca Cr Cu Fe Mg Ni Pb
1.137 1.095 0.000 0.000 1.453 0.005 0.000 0.000
0.000 0.693 0.000 0.043 0.000 0.301 0.613 0.000
0.000 2.036 0.000 1.580 0.481 0.080 0.022 0.137
0.175 0.000 0.230 0.178 0.000 2.304 0.076 0.000
0.000 1.481 0.963 0.000 1.087 0.353 0.000 0.373
S. Suryawanshi et al. / Science of the Total Environment 569–570 (2016) 46–52
49
Fig. 2. Percentage contribution of elements to different factors (Factor fingerprints).
in samples. Uncertainty is calculated using the following formula (Kim et al., 2005) uij ¼ p j xij þ MDL j Where,
3
ð2Þ
uij = uncertainty. pj = parameter proportional of uncertainty. xij = concentration of specie j in sample i. MDLj = Minimum Detection Limit. Positive Matrix Factorization is based on multivariate factor analysis which breaks down a matrix having species data into two matrices; factor contribution and factor profile. The factor profile
Fig. 3. Factor profiles.
50
S. Suryawanshi et al. / Science of the Total Environment 569–570 (2016) 46–52
Fig. 4. Source apportionment of indoor air pollutants.
needs interpretation by the user to identify the sources of pollutants which are contributing to the measured data by comparing the factor profile data with emissions and discharges from inventories/other sources. p
xij ¼ ∑ g ik f kj þ eij k¼1
ð3Þ
j - specie. Factor profile and contributions are obtained in PMF by minimising the objective function Q. " #2 p xij − ∑k¼1 g ij f kj Q ¼ ∑∑ uij i¼1 j¼1 n
m
ð4Þ
Where, xij - concentration of a species. eij - residual matrix. fkj - factor profile. gik - factor contributions. p - factor number. i - sample number.
Where, uij is uncertainty of specie j in sample i. There are two types of Q displayed by the model results, Q robust and Q true. Q true is calculated using all the points whereas Q robust is calculated by excluding the points which are not fit by the model, having uncertainty scaled residual greater than 4. The difference between Q robust and Q true shows the effect of inconsistencies from sources which are present during the sampling period (Norris and Duvall, 2014).
Fig. 5. Factor contributions (μg/m3) to the PM0.6 mass.
S. Suryawanshi et al. / Science of the Total Environment 569–570 (2016) 46–52
For choosing the number of factors, Q = i ∗ j is a good starting point. The model interpretation, Q values, the goodness of model fit and post PMF regression are all considered together to find the optimum number of factors. PMF require multiple iterations, it starts the search with a randomly generated factor profile. Since the starting point is random in nature there is no guarantee that PMF will give a global minimum, it may instead find a local minimum. So in order to find a global minimum PMF should be run multiple times iteratively. There can exist a multiple number of F (factor profile) and G (factor contribution) matrices giving the same minimum value of Q. This is known as rotational ambiguity. In order to check rotational ambiguity PMF has FPEAK tool. Five FPEAK runs could be performed in PMF for the current dataset. PMF accepts only non-negative values. It does not work with the missing data in data matrix. In case, if missing values are present then there are three options – remove the entire sample, remove the species or replace missing value with mean value. Measurements which are not relevant to the sources in study are discarded from the PMF analysis. Data below detection limit (DL) is replaced by DL/2 and its uncertainty is replaced by (5/6) DL. Species can also be categorised as bad, weak and strong based on signal to noise ratio and their uncertainties are also changed accordingly (Hopke, 2014). 3. Results and discussions PM0.6 and metal concentration varies a lot on different days as shown in Fig. 1. The maximum and minimum PM0.6 concentration was found out to be 330.79 μg/m3 and 17.59 μg/m3, respectively. The average PM0.6 concentration was 94.44 μg/m3. The statistical analysis of metal concentrations measured during the study period (monthly average) is shown in Table 1. PMF analysis was carried out on elemental concentrations taken over a period of time. The uncertainty values were calculated using Eq. (2) and values as depicted in Table 2. Several parameters were considered for obtaining the result. The missing values in the data were replaced by species median. Certain species were considered as bad species in PMF analysis and their uncertainty was replaced by a higher uncertainty value which was four times of the species-specific median. PMF was run for different number of factors. Q value decreased as the number of factors increased. Q theoretical for the data was approximately 784. Q (robust) for four and five factors was 1158.4 and 482.7, respectively. Pb was categorised as weak specie after analysing the base model run. F-peak analysis was also done to consider the rotational ambiguity. Five sources of indoor air pollutants were identified after PMF analysis (Table 3). G-Space plots in PMF showed that all the sources were independent and there was no correlation between them. Distribution of metal species in different sources shown in factor fingerprints plot (Fig. 2) gives the percentage contribution of metal species in different sources. It helps in the identification of major pollutant source releasing these metals. Factor 1 (Coal combustion) as shown in Fig. 3 had high concentration of Fe, Ba and Ca. A little amount of Mg was also present. These species are good indicator of coal combustion from households and from the power plants, brick kilns and industries present in and around the Kanpur city. According to a study Fe is present in the fly ash in coal in the form of Fe2O3 or Fe3O4 in about 2 to 26.8% (Cincinnati, 1969; Kang et al., 2011). Studies also shows release of CaO, Mg(OH)2 (Benson, 2006) and Ba (Kim and Kazonich, 1999; Kang et al., 2011) in coal combustion. Ba occurs in coal in different amount depending upon the type of coal (Zhao et al., 2014). Iron in coal is primarily associated with sulphur in minerals jarosite and pyrite, while other iron containing minerals like illite, ankerite and kaolinite also occurs in coal (Waanders et al., 2003). Calcium is generally present in coal as calcite (CaCO3) (Aladdin et al., 2013). A study showed that the ash deposits from coal fired power plant consists of CaO, MgO, Fe2O3 and other chemical compounds (Fernandez-Turiel et al., 2004). So this source indicates coal
51
combustion. The pollutants from industries are carried to the experimental site along with wind and enter the indoor environment through doors and windows. Factor 2 (Tobacco smoking) as shown in Fig. 3 had high concentration of Ca, Ni and significant amount of Mg, Cu. These pollutants are good indicators of tobacco smoking within indoor microenvironments. A study shows the effect of some important metals like Ca, Ni, Cu, Mg which are related to tobacco smoking (Chiba and Masironi, 1992). A study in China showed that both Chinese and imported cigarette papers had a significant amount of calcium and magnesium (Peng et al., 2011). Ni and Cu are toxic metals present in tobacco leaves (Bernhard et al., 2005). So this source indicates tobacco smoking. Factor 3 (Wall dust) as shown in Fig. 3 had high concentration of Ca, Cu and significant amount of Fe, Pb, Mg and Ni. Ca in the form of calcium carbonate is the main ingredient of wall putty used for finishing before painting the walls. According to studies, wall paints and construction materials are made up of metals like Pb, Zn, Cd, Cu, Mn, Ni, Co, Cr, Fe, etc. (Abagale and Campus, 2013; Mielke et al., 2001). So wall dust should be the source for this factor. Factor 4 (Soil particles) as shown in Fig. 3 had high concentration of Mg and significant amount of Cr, Cu, Ba and Ni. An USEPA study of 1983 shows the concentration of metals in the natural soil. Cr, Cu, Ba and Ni are present in the soil dust in approximately the same ratio as they are present in this factor (Lindsay, 1979). According to a study, soil consists of high concentration of Mg and a significant amount of Ba, Cu, Cr and Ni (Zeiner et al., 2015). So the source for factor 3 should be soil particles coming indoor along with wind or physical contact with people. Factor 5 (Wooden furniture/paper products) as shown in Fig. 3 had high concentration of Ca, Cr, Fe and significant amount of Mg and Pb. According to studies, wood has heavy metals like Ca, Mg, Al, Cd, Cr, Cu, Fe, Mn, Pb which may vary according to the climate and forest type (Arthur et al., 1999; Meier, 2013; Nicewicz and Szczepkowski, 2008). So the source of factor 5 could be the wooden furniture/paper products present indoor. The percentage share of different sources in PM0.6 (Fig. 4) shows that the maximum indoor pollution was caused due to wall dust (factor 3). Wall dust causes about 25.7% of indoor pollution. Wooden furniture/ paper products (factor 5) was the second highest pollutant source with a share of about 25.2%. Minimum indoor pollution was caused due to tobacco smoking (factor 2) with share of about 9.8%. Coal combustion (factor 1) and soil particles (factor 4) contributed about 21.8% and 17.5% respectively to indoor air pollution. Factor contributions of the sources were not constant and they changed with time (Fig. 5). This change in contribution of factors with time might be due to change in temperature, humidity and other influences.
4. Conclusions Sampling was done to collect PM0.6 concentrations at indoor sites in IIT Kanpur campus. Eight elements namely Ba, Ca, Cr, Cu, Fe, Mg, Ni and Pb were quantified using ICP-OES. A total of 98 samples were collected from November 2013 to September 2014 at suitable intervals. During this study it was found that there were large variations in PM0.6 concentrations with time PM0.6 concentration varies from minimum 17.59 μg/m3 to maximum 330.79 μg/m3. The average indoor PM0.6 concentration was 94.44 μg/m3. Source apportionment for the indoor air pollutants was done using PMF. PMF results gave source profile and source contributions of indoor air pollution. The analysis shows that five sources were responsible for the indoor pollution. The sources were independent of each other and there was no correlation between them. The source contribution showed variations with time. The five sources identified for indoor pollution were coal combustion (21.8%), tobacco smoking (9.8%), wall dust (25.7%), soil particles (17.5%) and wooden furniture/paper products (25.2%).
52
S. Suryawanshi et al. / Science of the Total Environment 569–570 (2016) 46–52
Acknowledgement This study was supported by the research fund from IIT Kanpur. References Abagale, S.A., Campus, N., 2013. Chemical Studies on the Composition of Natural Paint Pigment Materials from the Kassena-Nankana District of the Upper East region of Ghana. 3 pp. 13–23. Aladdin, M., Jian, J., Yang, Q., Chen, L.-C., Finkelman, R.B., Huang, X., 2013. Laboratory studies of the impact of calcite on in vitro and in vivo effects of coal dust: a potential preventive agent for coal workers' pneumoconiosis? Am. J. Ind. Med. 56, 292–299. Arthur, M.A., Siccama, T.G., Yanai, R.D., 1999. Calcium and magnesium in wood of northern hardwood forest species: relations to site characteristics. Can. J. For. Res. 29, 339–346. Benson, L.B., 2006. Use of Magnesium Hydroxide for Reduction of Plume Visibility in CoalFired Power Plants. Measurement. Bernhard, D., Rossmann, A., Wick, G., 2005. Metals in cigarette smoke. IUBMB Life 57, 805–809. Boss, C., Fredeen, K., 2004. Concepts, Instrumentation and Techniques in Inductively Coupled Plasma Optical Emission Spectrometry. PerkinElmer. Chakraborty, A., Gupta, T., 2010. Chemical Characterization and Source Apportionment of Submicron (PM1) Aerosol in Kanpur Region. Aerosol Air Qual. Res. 10, 433–445. Chauhan, Gupta, 2016. Development of a novel PM0.6 sampler (under review). Chiba, M., Masironi, R., 1992. Reviews/Analyses Toxic and trace elements in tobacco and tobacco smoke *. 70 pp. 269–275. Cincinnati, O., 1969. NEPIS document display. (URL:) http://nepis.epa.gov/Exe/ZyNET. exe/9100808I.TXT?ZyActionD=ZyDocument&Client=EPA&Index= Prior+to+1976&Docs=&Query=&Time=&EndTime=&SearchMethod= 1&TocRestrict=n&Toc=&TocEntry=&QField=&QFieldYear=&QFieldMonth= &QFieldDay=&IntQFieldOp=0&ExtQFieldOp=0&XmlQuery=&File=D% 3A\zyfiles\Index%20Data\70thru75\Txt\00000008\9100808I.txt&User= ANONYMOUS&Password=anonymous&SortMethod=h|-&MaximumDocuments= 1&FuzzyDegree=0&ImageQuality=r75g8/r75g8/x150y150g16/i425&Display= p|f&DefSeekPage=x&SearchBack=ZyActionL&Back=ZyActionS&BackDesc= Results%20page&MaximumPages=1&ZyEntry=1&SeekPage=x&ZyPURL (accessed 1.29.16). Devi, J.J., Gupta, T., Tripathi, S.N., Ujinwal, K., 2009. Assessment of Personal Exposure to Indoor and Outdoor Particulate Matter for Residents of an Academic Campus (IIT-Kanpur). Inhal. Toxicol. 21 (14), 1208–1222. Dherani, M., Pope, D., Mascarenhas, M., Smith, K.R., Weber, M., Bruce, N., 2008. Indoor air pollution from unprocessed solid fuel use and pneumonia risk in children aged under five years: a systematic review and meta-analysis. Bull. World Health Organ. 86, 390–394. Dockery, D.W., Pope 3rd, C.A., Xu, X., Spengler, J.D., Ware, J.H., Fay, M.E., Ferris, B.G.J., Speizer, F.E., 1993. An association between air pollution and mortality in six U.S. cities. N. Engl. J. Med. 329, 1753–1759. Fernandez-Turiel, J.L., Georgakopoulos, A., Gimeno, D., Papastergios, G., Kolovos, N., 2004. Ash deposition in a pulverized coal-fired power plant after high-calcium lignite combustion. Energy Fuel 18, 1512–1518. Gupta, T., Mandariya, A., 2013. Sources of Submicron Aerosol during Fog Dominated Wintertime at Kanpur. Environ. Sci. Pollut. Res. 20 (8), 5615–5629. Hopke, P.K., 2014. A Guide to Positive Matrix Factorization. pp. 1–16.
Kang, C.M., Gupta, T., Ruiz, P.A., Wolfson, J.M., Ferguson, S.T., Lawrence, J.E., Rohr, A.C., Goldleski, J., Koutrakis, P., 2011. Aged particles derived from emissions of coal-fired power plants: The TERESA field results. Inhal. Toxicol. 23, 11–30. Kim, A.G., Kazonich, G., 1999. Mass Release of Trace Elements from Coal Combustion ByProducts. Kim, E., Hopke, P.K., Qin, Y., 2005. Estimation of organic carbon blank values and error structures of the speciation trends network data for source apportionment. J. Air Waste Manage. Assoc. 55, 1190–1199. Klepeis, N.E., Nelson, W.C., Ott, W.R., Robinson, J.P., Tsang, a.M., Switzer, P., Behar, J.V., Hern, S.C., Engelmann, W.H., 2001. The National Human Activity Pattern Survey (NHAPS): a resource for assessing exposure to environmental pollutants. J. Expo. Anal. Environ. Epidemiol. 11, 231–252. Koenig, J.Q., Mar, T.F., Allen, R.W., Jansen, K., Lumley, T., Sullivan, J.H., Trenga, C.A., Larson, T.V., Liu, L.J.S., 2005. Pulmonary effects of indoor- and outdoor-generated particles in children with asthma. Environ. Health Perspect. 113, 499–503. Lighty, J.S., Veranth, J.M., Sarofim, A.F., 2000. Combustion aerosols: factors governing their size and composition and implications to human health. J. Air Waste Manage. Assoc. 50, 1522–1565. Lindsay, W.L., 1979. Chemical Equilibria in Soils. Wiley-Interscience Publication, Wiley. Meier, S., 2013. Elemental analysis of wood fuels NYSERDA's promise to New Yorkers : NYSERDA provides resources, expertise and objective information so New Yorkers can make confident, informed energy decisions. NYSERDA Rep. 13–13. Mielke, H.W., Powell, E.T., Shah, A., Gonzales, C.R., Mielke, P.W., 2001. Multiple metal contamination from house paints: consequences of power sanding and paint scraping in New Orleans. Environ. Health Perspect. 109, 973–978. Nicewicz, D., Szczepkowski, A., 2008. The content of heavy metals in the wood of healthy and dying beech trees (Fagus sylvatica L.). Acta Sci. Pol. Silv. Colendar. Rat. Ind. Lignar. 7 (4), 35–44. Norris, G., Duvall, R., 2014. EPA Positive M atrix Factorization (PM F) 5.0 Fundamentals and User Guide. Peng, W., Jun, S., Xiangmin, L., Lijun, Z., Li, L., 2011. Determination of calcium and magnesium in cigarette paper with autoanalyzer. Tob. Sci. Technol. 52–55. Perez-Padilla, R., Schilmann, A., Riojas-Rodriguez, H., 2010. Respiratory health effects of indoor air pollution. Int. J. Tuberc. Lung Dis. 14, 1079–1086. Rohra, H., Taneja, A., 2016. Indoor air quality scenario in India—an outline of household fuel combustion. Atmos. Environ. 129, 243–255. Smith, K., 1993. Fuel combustion, air pollution exposure, and health: the situation in developing countries. Annu. Rev. Energy Environ. 18, 529–566. Smith, K.R., 2000. National Burden of Disease in India From Indoor air Pollution. Smith, K.R., Samet, J.M., Romieu, I., Bruce, N., 2000. Indoor air pollution in developing countries and acute lower respiratory infections in children. Thorax 55, 518–532. Uzoigwe, J.C., Prum, T., Bresnahan, E., Garelnabi, M., 2013. The emerging role of outdoor and indoor air pollution in cardiovascular disease. N. Am. J. Med. Sci. 5, 445–453. Waanders, F.B., Vinken, E., Mans, A., Mulaba-Bafubiandi, A.F., 2003. Iron minerals in coal, weathered coal and coal ash - SEM and Mossbauer results. Hyperfine Interact. 148149, 21–29. Zeiner, M., Juranovic Cindric, I., Pozgaj, M., Pirkl, R., Silic, T., Stingeder, G., 2015. Influence of soil composition on the major, minor and trace metal content of Velebit biomedical plants. J. Pharm. Biomed. Anal. 106, 153–158. Zhao, C.L., Sun, Y.Z., Xiao, L., Qin, S.J., Wang, J.X., Duan, D.J., 2014. The occurrence of barium in a Jurassic coal in the Huangling 2 mine, Ordos Basin, northern China. Fuel 128, 428–432.