Chemical Source Profiles of Emissions Resulting from Industrial and Domestic Burning Activities in India

A study has been performed to develop PM2.5 (particles with aerodynamic diameters ≤ 2.5) chemically speciated source profiles of different industrial and domestic burning practices in India. A total of fifty-five PM2.5 samples have been collected in emissions resulting from (1) industrial furnaces, (2) household fuels, (3) municipal solid waste burning, and (4) welding workshop burning practices, and categorized for eleven subtypes of sources. The collected samples were subjected to chemical analysis for twenty-one elemental (Al, As, Ca, Cd, Co, Cr, Cu, Fe, Hg, K, Mg, Mn, Mo, Na, Ni, Pb, S, Sb, Se, V, Zn), nine ionic (Na, K, Mg, Ca, NH4, Cl, F, NO3, SO4), OC, and EC source indicator species using atomic absorption spectrometry, ion chromatography and carbon analysis (thermal/optical transmittance method), respectively. The carbonaceous fraction was most abundant in household fuel burning emissions (47.6 ± 7.45% to 65.92 ± 13.13%). The ionic/elemental ratios of major inorganic constituents (Ca/Ca, Mg/Mg and Na/Na) have been identified to describe the PM2.5 emissions from combustion or re-suspension dusts during industrial activities. Brick Kiln processes (BKP) have been identified as the major emitter of the highest number of toxic species (Cd, Co, Mo, Sb and V), followed by steel rerolling mills (Hg and Pb) and steel processing industries (As, Ni). The source marker calculations also confirmed that K, Mn, and As are good markers for biomass burning, metallurgical industrial emission, and coal burning, respectively, similar to the findings in previous studies.


INTRODUCTION
The source apportionment of PM fractions increaseswith different trends using receptor models, mostly chemical mass balance (CMB), to develop pollution control and mitigation strategies worldwide (Watson et al., 2002;Samara, 2005;Khan et al., 2010;Kong et al., 2010).Receptor models that derive profiles from ambient/indoor measurements requires systematic chemically speciated emission profiles from prominent sources that were possible to effect pollutant concentration at receptor for verification (Hopke, 1999;Watson et al., 2001Watson et al., , 2002;;Brook et al., 2003;Gupta et al., 2007).These source profiles are the fractional mass (abundances ± uncertainty) of measured chemical species relative to primary PM mass of source emissions (Watson et al., 2001) and one of the most important parameters (Pant and Harrison, 2012) to: 1) create chemically speciated emission inventories (Cass and McRae, 1983;Kuykendal et al., 1990;Chow et al., 2004), 2) apportion receptor concentrations to source (Watson et al., 1984(Watson et al., , 1990(Watson et al., , 1991(Watson et al., , 2001) ) and 3) estimate toxic and hazardous pollutant emissions (Chow et al., 2004).Chemical abundance in most of earlier source profiles is accompanied by an uncertainty/standard deviation value that intends to represent the errors/variability of that abundance resulting from differences among separate emitters and between samples taken same/different times from the same emitters; which is essential to CMB runs (Watson et al., 1994(Watson et al., , 2001;;Chow et al., 2003;Ho et al., 2003;Chow et al., 2004;Tsai et al., 2007).
The current study presented PM 2.5 chemical source profiles of emissions resulting from 07 different industrial processes including arc-welding workshops and 04 domestic burning practices involved with household cooking activities and municipal solid waste management in India.These PM 2.5 chemical source profiles were developed with the objectives to meet the requirement of location specific and latest source profiles that could be applied for chemical mass balance receptor modelling studies; and to update previous source profiles.

METHODOLOGY
The development of PM 2.5 chemical source profiles of selected burning practices have been carried out as a part of a comprehensive source apportionment study of indoor/ outdoor PM 2.5 in a dense urban-industrial zone of India (Balakrishna and Pervez, 2009;Pervez et al., 2012) by following a real-world pooled sampling plan using purposeful study design (Gilbert, 1987).Sampling of PM 2.5 was conducted at four different types of combustion sources, mainly observed in urban areas of Chhattisgarh, India: (1) Municipal waste burning, (2) Household fuel burning (3 sub-types), (3) Mineral based coal fired industries (6 subtypes) and (4) Fabrication workshops (Table 1) (Balakrishna andPervez, 2009, 2011;Pervez et al., 2012).

Description of Study Area and Source Characteristics
Raipur-Bhilai, major industrial cities of Chhattisgarh, India located in global scale of 21°14′22.7′′N,81°38.1′′Eand 21°11′0′′N, 81°23′6′′E respectively, having population 1,635,784 (Census, 2011), is known for most dense heavy industrial zone composed of mainly iron processing, thermal power generation and cement production activities.Identification of PM 2.5 emission sources were based on previous reported air monitoring studies (Dubey and Pervez, 2008;Balakrishna et al., 2011;Pervez et al., 2012), layout map, and a survey of current burning practices involved with industrial and domestic activities.About 1200 tonnes of municipal solid waste (MSW) is generated every day in Raipur-Bhilai, region and about 65% of the MSW (about 650 tonne) was disposed-off using open burning procedure on daily basis.Pervez et al. (2012) reported that all populations  1 and Fig. 1.

Sampling
Two different sampling methodologies have been adopted according to nature and characteristics of different burning practices: real-world in-plume and re-suspension sampling (Chow et al., 2004;Patil et al., 2013).In case of industrial combustion processes, stack emitted bag filter house dust samples were re-suspended tocollectPM 2.5 fractions using standard procedures reported elsewhere (Chow et al., 2004;Gadkari and Pervez, 2007).In case of open burning sources related to household fuel burning (RSFS, RKS, RLPGS), outdoor municipal waste burning practices (MSWB), arcwelding workshops (EAW) and brick kilns (BKP), PM 2.5 was sampled from smoke plume.All these open burning sources does not have stacks and PM 2.5 impactors were positioned in smoke plume.Before in-plume sampling, background PM 2.5 were measured for subtraction from the in-plume concentration (Chakrabarty et al., 2013;Dewangan et al., 2013).In case of PM 2.5 sampling using chamber resuspension procedures, sampling duration was optimized according to standard filter loading conditions (Chow et al., 2003;DRI, 2011).In case of in-plume sampling for open burning sources, PM 2.5 sampling event was conducted in 3-4 episodes to cover whole burning processes (Chow et al., 2003;DRI, 2011;Chakrabarty et al., 2013).PM 2.5 has been collected on quartz fiber filters (QFF) (1851-047, Whatman, UK) using Parallel operation of five PM 2.5 samplers (MINIVOL, Ver.4.2, Model AirMatrics) in each source site at average flow rate of 5 L/min.Filter selection, preparation, calibration, installation, transportation, preservation, weighing measurements and field blanks were conducted by following the quality control and quality assurance described in air sampling protocol reported elsewhere (USEPA, 1999a, b;CPCB 2008;DRI, 2011;Patil et al., 2013).Details of PM 2.5 source sites, their location frequencies in study region and sampling methodology used has been described in Table 1.
As far as OC and EC analysis is concern, thermal optical transmittance (TOT) method presented by National Institute for Occupational Safety and Health-NIOSH-5040 protocol (NIOSH, 1999) using semi-continuous thermal/optical carbon analyzer (Sunset Laboratory, Model 4L, USA) (Birch and Cary, 1996;Schauer, et al., 2003;Pipal et al., 2014) has been adopted.Concentration of selected chemical species measured in PM 2.5 were corrected by subtracting them from those found in field blanks; followed by subtraction from those found in background PM 2.5 .The background corrected concentrations along with their uncertainties and in-plume/ background ratio (for those measured in smoke plume) have been presented in Table 2 and Table 3. Uncertainty of source profile abundances estimated as the standard deviation of the average from five source tests (Watson et al., 2001;Watson and Chow, 2007).Species concentrations in laboratory blank and field blanks have also been presented in Table S1 (supporting information).Chemical species abundances in four different ranges of percent by weight along with comparison with reported values have been presented in Table 4 and Table 5. Components of crustal origin, ionics, trace elements and carbonaceous matter found in selected source profiles have been presented in Fig. 2. Source markers of selected source emitted PM 2.5 has been shown in Table 6.

Description of PM 2.5 Chemical Source Profiles
A source profile comprised offivefractional abundances of individual chemical species with respect to total PM mass and individual uncertainty or standard deviation value for particular chemical species (Watson et al., 2001).Chemical profiles have shown discernible pattern in relative strengths of selected species in PM 2.5 emissions resulting from industrial and domestic burning practices.Profile-wise measured mass accounted were: 49.30 ± 3.35%.

Municipal Solid Waste Burning (MSWB)
Municipal refuse open burning is one of the major contributors to the air pollution in Indian cities (Patil et al., 2013).PM 2.5 emissions from MSWB is prominently loaded with TC (84% of measured mass) having OC/TC ratio of 0.94 and potassium (5.14% of measured mass) having K + /K ratio of 0.77; comparable to those reported for vegetative and biomass burning (Watson et al., 1994(Watson et al., , 2001;;Chow et al., 2004).Abundance of other ions (Na + , NH 4 + , Ca 2+ ) were significantly higher (> 1% of measured mass) than species related to crustal origin and trace elements.

Residential Fuel Stoves
Residential cooking stoves are mostly uses three different types of fuel/fuel mixtures for cooking purposes in India: solid fuel mixture (3:1:1 ratio of Coal, wood, dung cakes), kerosene and LPG.Highest abundance of TC (75-87% of measured mass) in PM 2.5 emissions has been observed.In addition, RSFS has shown OC/EC ratio of 6.9; two and  Chow et al., 2004 As, Mg, Cd, Co, Cr, Cu, Hg, Mn, Mo, Ni, Pb, Sb,  , OC, EC Watson et al., 2001 Cd, Co, Cr, Cu, Hg, Mn, Mo, Ni, Pb, Sb, Se, V, Zn As, Cl, K, Mg, Na,NO    a, b, d  Mn, Zn, Pb, Cd, Cu, As, Hg Industry a, c and Coal Combustion b V, Ni, SO 4 2-and Se, As, Cr, Co, Cu, Al * Reported sources and their markers in italic style with underline for comparison.a Watson et al., 2008; b Mitra et al., 2002; c Viana et al., 2008; d Viana et al., 2006.five-fold higher compared to RKS and RLPGS, respectively.On considering the biogenic markers (K + and NH 4 + ), higher values of K + /K ratio (0.89-0.97) and NH 4 + abundance compared to those found in crustal origin and industrial emitted PM 2.5 is attributed to the fact that biogenic sources are dominating in PM 2.5 emissions from household solid fuel burning practices involved with cooking purposes.These values are similar to earlier reports on emissions of biogenic (Watson et al., 1994(Watson et al., , 2001;;Chow et al., 2004) and solid fossil fuel combustion (Watson et al., 2001;Kong et al., 2011).About 4-13 times higher K + emissions has been observed in case of RSFS compared to RKS and RLPGS, respectively.But NH 4 + was found 2-16 times higher in emissions resulting from kerosene stoves (RKS) compared to RLPGS and RSFS, respectively.As far as trace species is concern, sulphur constituents (S, SO 4 2-) along with Se and seven toxic species (As, Cd, Hg, Ni, Pb, V and Zn) were found multi-fold higher in RKS compared to RSFS and RLPGS; similar to earlier reported levels for kerosene stoves (Patil et al., 2013).Other anions (NO 3 -, F -and Cl -) have shown higher presence in PM 2.5 emissions resulting from RSFS compared to RKS and RLPGS.Most of markers of crustal origin are found < 0.1% in all types of household fuel burning emissions.

Coal Fired Mineral Based Industries
Six different process based industrial emissions, mostly dominating in study region, have been chosen for the development of PM 2.5 chemical source profiles.Out of six chosen industries, only BKP has shown > 50% load of carbonaceous matter (TC) in their PM 2.5 emissions.Other industrial PM 2.5 emissions have shown lesser load of TC in the ranges of 8.03-35.53%.PM 2.5 emissions from two industrial combustion processes (SPI and BKP) have shown higher load of EC compared to OC with EC/OC ratio of 3.25 and 1.55, respectively.Significantly lower measured mass in case of CPI, FEMNI and CTPP is might be due to higher unaccounted silica load in PM 2.5 mass.As far as iron particle load in PM 2.5 is concern, sponge iron processing (direct reduction iron processing) (SPI) have shown 2-34 times higher load compared to other industrial processes included in the study.In case of other major constituents like Al is found highest in CTPP (15.5% of measured mass) and Ca is found highest in CPI (39.6% of measured mass).Water soluble/total ratio (Ca 2+ /Ca, Mg 2+ /Mg and Na + /Na) describes the pre-dominance of emissions resulting from either dust re-suspension during process handling or combustion activities involved with mineral based coalfired industries (Volkovik, 1983;Watson et al., 1994).The lower values of these ratios in PM 2.5 emissions from CPI (0.13, 0.01 and 0.33, respectively), FEMNI (0.12, 0.25 and 0.58, respectively) and SRM (0.42, 0.31 and 0.18, respectively) attributed the higher abundance from crustal origin, whereas higher values for SPI (0.71, 0.55 and 0.61, respectively), CTPP (0.69, 0.73 and 0.68, respectively) and BKP (0.69, 0.28 and 0.81, respectively) indicated the predominance of emissions from combustion processes.CTPP, SPI and BKP have shown higher enrichment of biogenic potassium (K + ) with K + /K ratio of 0.73, 0.64 and 0.62, respectively; contrast to CPI and SRM along with earlier reported values for coal fired steel industries (Watson et al., 2001;CPCB, 2008;USEPA, 2013).This might be due to use of poor quality coals in CTPP and SPI and biomass (dung cakes with coals) in brick kilns .Order of variation in association of anions with PM 2.5 emissions from selected source sites is evaluated to be: NO 3-> F -> Cl -> SO 4 2-.onother hand, uniformity in occurrence of sulphur group species (S, SO 4 2-and Se) across the selected industrial sites is attributed to similar source origin of coal combustion (Volkovik, 1983).Trace elements have also shown significant variation in their relative abundances in PM 2.5 emitted from different sources; justified their inclusion in the development of PM 2.5 chemical source profiles for emissions resulting from combustion processes involved with different mineral based coal-fired industries.Brick Kiln processes (BKP) has been identified as the major emitter of highest number of toxic species (Cd, Co, Mo, Sb and V) followed by SRM (Hg and Pb) and SPI (As, Ni).Moderate combustion temperature and poor quality of coals and other combustion materials (dung cakes etc.) are responsible for higher emission of toxic species (Vollkovic, 1983).
In addition to these industrial sources, open fabrication workshops having arc-welding activities (EAW) is also reported to be the significant contributor of outdoor PM 2.5 due to their profuse locations within the study region (Pervez et al., 2005).It has been observed that EAW is contributing mainly active iron and manganese particles.Pb, As, Cu and F -were observed to be found above 0.1% compared to other toxic species similar as earlier reported profiles of arc-welding workshops (Swamy et al., 1994;CPCB, 2008).

Comparison of Developed Source Profiles with Reported Profiles of National and International Origin
To compare the developed PM 2.5 chemical source profiles in this study with previous reported profiles by Chow et al. (2004), Watson et al. (2001), the Central Pollution Control Board, India (2008) and Speciate 4.0 (USEPA).All chemical species of each of selected profiles are grouped in four percentage fractional ranges (Chow et al., 2003(Chow et al., , 2004) ) and summarized in Table 4 and Table 5.The major markers (OC, K + , NH 4 + ) of emissions resulting from MSWB were observed to be found in different levels in developed and reported profiles.OC was found > 10% in present and CPCB profiles, but that was observed to be within 1-10% in Speciate 4.0 profiles.K + is found in different fractional range between present and CPCB profiles with higher abundance (1-10%) in presented profile; confirm the strong variation in biogenic matter content in MSWB.Additionally, the different abundance of K + in MSWB emitted PM 2.5 between present and CPCB profiles might be due to different sampling methods; CPCB adopted laboratory scale study, whereas real-world sampling on open air burning was adopted in present study.Apart from major markers, other species of crustal origin were found in similar fractional ranges in MSWB profiles developed in present study and CPCB database as well.In case of PM 2.5 profiles of emissions resulting from residential fuel burning (RSFS, RKS and RLPGS), carbonaceous matter was found > 10% across all the comparative profiles, but EC was found > 10% only in presented profiles.As far as anions and cations of fuel combustion markers (K + , NH 4 + , SO 4 2-, Cl -, NO 3 -) is concern, most of them found higher (1-10%) in presented profiles compared to databases of CPCB and Speciate 4.0; most of them found higher in emissions of kerosene stoves (RKS).Other species of crustal origin and trace elements have shown more than 50% agreement in their occurrence in similar fractional groups across the selected databases for comparison.
On comparing PM 2.5 chemical profiles of emissions resulting from selected industrial sites with CPCB and Speciate 4.0, different observations were obtained for different industrial sites.Sponge-iron industrial emissions (SPI) and Ferro-manganese industry (FEMNI) were not included in CPCB and Speciate 4.0.In case of CPI and CTPP, major fractional group (> 10%) has shown different inclusion of species across the comparative databases.BKP has shown inverse distribution of OC and Ca between two fractional groups (> 10% and 1-10%) on comparing the presented profile with CPCB profile.On contrary, EAW has shown similar major marker (Fe) in both comparative source profile databases.The element Ca for CPI, OC and EC for SRM & BKP, Fe for EAW, EC for SPI and OC for FEMNI were accounted for > 10% abundances.The water soluble ions and crustal element distributed between 0.1-1% and 1-10% ranges.The trace elements were contributed < 0.1% relative abundances except As and S for CTPP with range 0.1-1%.
Overall species abundance of source profiles in defined fractional groups is in ~60-80% agreement with reported source profile databases of CPCB and in ~35-50% agreement with USEPA Speciate 4.0.

Mass Closure Analysis
Mass reconstruction of PM 2.5 has been carried out using revised IMPROVE18 mass closure (material balance) equation (Ho et al., 2003, Chow et al., 2012); estimate the unmeasured oxides and compared with the total gravimetric measured mass for the quality assurance (Watson et al., 2012)

Ion Balance Calculation
The ionic balance calculation has been performed to confirm the acid-base property of PM 2.5 fractions emitted by different burning practices.Conversion of ion mass concentrations into micro equivalents was performed to calculate the cation/anion balance of PM 2.5 (Cao et al., 2005;Zhang et al., 2011;Tao et al., 2013).The cation and anion micro equivalents of particles were calculated as follows: C (Cation micro equivalents/m 3 ) = Na + /23 + NH 4 + /18 + K + /39 + Mg 2+ /12 + Ca 2+ /20 (2) A (Anion micro equivalents/m 3 ) = F -/19 + Cl -/35.5 + NO 3 -/ 62 + SO 4 2-/48 (3) The well balanced anion/cation (A/C) ratio must be 1.The value higher than 1, indicates the acidic nature of the particle (Kerminen et al., 2001) whereas slightly lower than 1, indicate contribution of unmeasured CO 3 2-ion, and very low A/C ratio indicates the basic nature of particle (Cao et al., 2005;Shen et al., 2007Shen et al., , 2009)).A/C ratios of PM 2.5 emissions resulting from domestic and industrial burning practices have been accounted; ranges from 0.18-0.79and 0.19-0.74,respectively and confirm that all source emitted PM 2.5 samples were basic in nature.

Source Markers
Source markers of particles are mostly described by specific size distribution, specific suite of elements and specific ratios of compounds, elements or isotopes (Mitraet al., 2002).The relative source indicator species were evaluated for all eleven sources grouped in domestic and industrial burning practices.For the calculation, following formula was applied to define the indicatory species for specific source emitted PM 2.5 fraction (Yang et al., 2002;Kong et al., 2011): , min ( / ) ( / ) where: X i was the i th individual species concentration; (X i /ΣX) j was the quotient of i th individual species ij divided by the summation of 32 species concentrations of emission source j; (X i /ΣX) min was the quotient of i th individual species divided by the summation of 32 species concentrations which were the minimum for all emission sources (Yang et al., 2002;Chen et al., 2003).A normalization procedure has been applied according to Mitra et al. (2002) and Kong et al. (2011) to minimize the effect of physical parameters.Normalized individual species concentration was used by dividing the i th individual species ij concentration to the sum of i th individual concentration for all the source profiles (Kong et al., 2011).The top six chemical species with highest ratio values for total relative source profiles has been use as relative source indicatory chemical species describe in detail on Table 6 and compare with earlier reports (Kong et al., 2011).Water soluble K + , marker of vegetative and biomass burning sources (Watson et al., 2002(Watson et al., , 2008)), F - and Ca 2+ were evaluated to be the similar source markers of PM 2.5 emissions from burning practices involved with MSWB and RSFS; whereas RKS and RLPGS have shown distinct source markers with major marker groups of (F - and V) and (S and Mo), respectively.Arsenic (As) was found common source marker of PM 2.5 emissions from CTPP, SPI, and BKP; one of prominent trace element marker of emissions resulting from coal burning (Mitra et al., 2002).Mn was calculated as common source marker PM 2.5 emissions from FEMNI and EAW.Distinct observations of source markers of PM 2.5 emissions from selected source sites compared to those reported earlier (Mitra et al., 2002;Viana et al., 2006;Watson et al., 2008) is attributed to the importance of development of region specific source profiles to obtained precise results of receptor modeling.

CONCLUSIONS
The differences with earlier reported/developed similar characteristics profiles created demand of additional and more precise source profiles that represent a study area.In this channel, the present study is an important work in development of source profile database in India.The eleven important stationary sources profiles for PM 2.5 fraction are reported in this paper.These profiles are comprised of 21 element, 9 water soluble ion, and carbonaceous fractions by following the standard protocol of chemical analysis and data validation.The carbonaceous fractions are most abundant with different OC/TC ratio; ranges from ~0.58-0.94 in PM 2.5 from selected source emissions.Trace metals were found significantly higher in PM 2.5 emissions from burning practices involved with household cooking activities and municipal solid waste management practices, compared to crustal origin.Observation of different relative enrichment of defined chemical components (Fig. 2) in PM 2.5 emissions from selected industrial burning sources might be due to use of different raw materials, and combustion temperature and conditions involved with industrial processes.The developed profiles comparatively much similar in > 10% and < 0.1% abundant species with earlier reported profiles for similar sources.The K + , Mn, and As were found and source marker for biomass burning, metallurgical industrial emission, and coal burning respectively, shown good agreement with National CPCB, 2008 and global USEPA speciate database also previous reported profiles.
These profiles require update, up gradation and addition of new sources with the sufficient interval of time to better represent changes in characteristics of sources of burning practices in India.

Table 1 .
Description of domestic and industrial source characterization, material used in burning practices and sampling method.
of this area use three categorized stoves (based on fuels) for household burning purposes, namely-Liquid petroleum gas (LPG) stoves, Kerosene stoves and stoves with conventional solid fuels.The census of India (2011) describes the household statistical figures of different type of fuel use for cooking purposes: 25% LPG stoves, 30% kerosene stoves and 45% stoves with conventional solid fuel in Central India.As far as different industrial processes carrying out in the study region are concern, about 1351 industries are currently existing in the study region; out of that heavy, medium and small scale industries numbers are 114, 295 and 942, respectively (DoCI, 2012).Nearly 300 major and medium units of iron processing with consumption of 20 Million Tonne (MT) of iron ore/scrap steels per year, ~50 small, medium and major units of coal-burning power generation with coal consumption of 19.03 MT/yr, 15 units of cement production with lime stones/slag/gypsum consumption of 26 MT/yr occurred in previous years.A total of 32.11 MT/yr coals are consumed in production of power generation, steel processes and cement production in the study area.Accordingly, eleven different types of combustion/burning practices involved with industrial, household and outdoor activities have been chosen for the development of PM 2.5 chemical source profiles.Details of source types, justification of their selections, combustion material used have been described in Table

Table 2 .
PM 2.5 composite sources profiles (weight percent by mass) of emissions from residential fuel and municipal waste burning practices.

Table 3 .
PM 2.5 composite sources profiles (weight percent by mass) of emissions from industrial burning practices.

Table 4 .
Summary of chemical species abundance in PM 2.5 emissions from domestic burning practices and their comparison with nationally and internationally reported

Table 5 .
Summary of chemical species abundance in PM 2.5 emissions from industrial burning practices and their comparison with nationally and internationally reported

Table 6 .
Source signatures of domestic and industrial burning practices in India.
-, As, Mg 2+ , Ca 2+ , Cr, K + CPI S, Cr, Cu, Al, Mo, Mg RKS Pb, Cd, Sb, F -, Se, V FEMNI Cr, Mo, Mg 2+ , Mn, NO 3 -, Cu RLPGS Sb, Cd, Pb, S, Mo, Se SPI Cr, As, Mg 2+ , Cu, EC, Cl - Residential fuel combustion a* OC, EC, K + , Cl - CTPP As, Cr, S, F -, NO 3 -, Al Municipal waste burning SRM Pb, Mg 2+ , NO 3 -, EC, S, Cl MSWB F-, Co, Cd, Ca 2+ , Na + , K + EAW Mn, Cr, K + , Cd, Pb, Mg 2+ Waste burning a OC, EC, K + , As, Pb, Zn BKP As, Cd, Mo, EC, NO 3 -, Sb Metallurgy Two major crustal elements, silica and titanium have not been included in mass closure study; resulting in lower estimated values for crustal fraction in the overall mass closure.In case of RKS and RLPGS, PM 2.5 mass reconstruction has been achieved 107.27% and 102.08%, respectively due to negligible abundances of unmeasured silica, titanium and higher abundance value of OC.Domestic solid material burning practices (RSFS and MSWB) have shown relatively lower mass closure values (97.47% and 79.36%, respectively).PM 2.5 Mass closure results for industrial sources have shown close equivalence between the gravimetric mass and reconstructed mass.The highest and lowest mass closure values have been found in case of EAW (98.84%) and CTPP (45.65%), respectively.Mass closure results of PM 2.5 emissions, resulting from other industrial source sites, were achieved 77.13%, 62.11%, 66.32%, 68.12% and 85.08% for BKP, FEMNI, SPI, CPI, and SRM, respectively.No overestimation of mass closure result has been occurred in developed profiles.