Release of Harmful Air Pollutants from Open Burning of Domestic Municipal Solid Wastes in a Metropolitan Area of Korea

Open burning of solid wastes, a potential nonpoint emission source, has recently become a topic of interests, particularly in the metropolitan area of Seoul, Korea. To estimate the effects of irregular open burning on local air quality, we evaluated the emission levels of harmful substances from test combustion of individual types of domestic municipal solid waste (MSW), including paper, wood, and plastics. The emission factors of PM10, PM2.5, PAHs, and heavy metals obtained from laboratory tests differed depending on the combustion material. A brief survey of residents and local government officials showed that more than 10.6% of homes in the metropolitan area have eliminated waste by irregular burning. Based on the public survey in the test area, the average annual emissions of hazardous materials from frequent open burning was estimated to be 71 tons for PM10, 46.6 tons for PM2.5, 914 kg for heavy metals, and 67 kg for PAHs. Open burning creates nearly 0.44% of regional air pollution from PM10.


INTRODUCTION
Open burning is the disposal behavior of municipal refuse, construction scraps, agricultural residues, etc., by uncontrolled combustion in an open yard or in barrels.These conditions result in incomplete combustion and release harmful air pollutants, such as fine particulates, PAHs, heavy metals, and dioxins.Open burning is common in urban and rural areas that have volume-rate disposal policies, which charge for disposal of household wastes based on the amount disposed.Most previous research regarding open burning have primarily dealt with the emission characteristics of harmful elements for rural areas (Park et al., 2005;Moon, 2007).Kakareka et al. (2003) emphasized the significant risk of open burning by analyzing the PAHs generated from burning post-cultivation agriculture debris.Ortiz de Zarate, et al. (2005) studied the emission of harmful elements by domestic open burning in Spain.Lemieux et al. (2004) classified 20 burning activities, in addition to agricultural burning, into characteristic types and closely observed potential toxic organic emissions including VOCs, PAHs, and dioxins.Park et al. (2005) evaluated carbon monoxide and carbon dioxide emissions by combusting biomass, including maple and ginkgo leaves, wild bushes, and pine tree leaves.Moon (2007) burned a mixture of agricultural residues in a test incineration chamber designed by the United States Environmental Protection Agency (U.S. EPA) and estimated the overall emissions and the emission factors of dioxin and PAHs.Kim (2008) found that a large quantity of harmful substances could be emitted from combusting municipal solid waste (MSW) and agricultural waste.Lee et al. (2006) investigated the handling of MSW disposed in small rural counties and analyzed incineration residue, with a focus on seven heavy metals in open yards.
Most of the aforementioned research relied primarily on rural waste burning.Little work has assessed the contribution of open burning to total atmospheric emissions in metropolitan areas, which is particularly concerning for those with a population density as high as Seoul, one of the largest capital cities in the world.The contribution that small-scale, widely-distributed open burning of waste piles makes to urban air pollution should be determined.Therefore, we evaluated emission factors and the quantities of fine particulates, heavy metals, and PAHs released from burning domestic MSW, including paper, wood, and plastics, in a laboratory-scale incineration chamber.

Test Wastes and Incineration Chamber
The test waste samples were selected based on the literature and a simple survey of metropolitan area residents.The most commonly burned municipal materials were waste paper, used plastics, waste wood, and household solid waste (MSW).Test MSW samples were randomly collected from a local waste depository in Y-city, which represents combined urban and rural customs near Seoul (Table 1).
Based on a study by Ortiz de Zarate et al. (2000), the test incineration chamber was designed with 2 mm thick stainless steel in the following dimensions: 500 × 300 × 400 mm (Fig. 1).A ceramic insulator inside the chamber prevents thermal transfer, and seven round holes on the lower portion of the three faces allow for natural air inflow.The web on the bottom aids in smooth mixing of air and combustibles.The flue gas exits through a 100-mm diameter duct with a hole at 1.5 m to sample dust, collect gas, and measure flow rate.

Test Burning and Sample Analysis
The test burning was carried out with 1 kg of each combustible in the laboratory.Combustion was considered complete upon reaching an oxygen concentration of greater than 20.5% in the flue gas.Combustion of a sample of MSW usually took about 1 hour.PM 10 and PM 2.5 were sampled separately by an isokinetic cyclone collector at an average rate of 15 L/min (Fig. 2).Separated particulate matter was deposited on quartz filters (Whatman International Ltd., QMA, UK).Since the highest emissions usually occur when the base of the combustibles is ignited, particulate matter from the flue gas was sampled beginning immediately after ignition until the end of combustion.No fresh waste remained in the bottom ash after test incinerations.The  oxygen concentration of the flue gas and temperature were analyzed in real-time using a flue gas analyzer (KM-9106, Kane-May, Korea).In order to increase the precision and accuracy of the data, each test burning was repeated at least 10 times.
Heavy metals in the flue gas (Pb, Ni, Cu, Cd, Cr, and Zn) were analyzed by inductively coupled plasma mass spectrometry (ICP-MS; Direct Reading Echelle ICP, LEEMAN) with a microwave pretreatment.Heavy metal elements collected in quartz filters were extracted using a refrigerated mixture of nitric acid and hydrochloric acid.The precision and accuracy of ICP-MS analysis were assessed by evaluating the coefficient of variation (CV) and relative error (RE), which is relative to the real concentration (Oh, 2008).While the CV for metal elements varied from 1.23 for Zn to 3.09 for Ni, RE varied from 0.43 for Ni to 2.25 for Cr.Thus the analysis of metal elements using ICP-MS was fairly reliable.Particulate PAHs were analyzed by gas chromatography/mass spectrometry (GC/MS) through Soxhlet extraction, Young Nam University, approved by ministry of Environment in Korea (U.S. EPA, 2008).
The temporal evolution of the combustibles was assessed using a smoke chamber, which was designed according to ISO 5659-2 (Determination of Optical Density by a Single-Chamber Test) and ASTM E 662 (Standard Test Method for Specific Optical Density of Smoke generated by Solid Materials).The smoke density was determined by relative light penetration of smoke immediately after combustion.Test aliquots of 7.5 × 7.5 × 1.5 cm were ignited in the chamber and the Ds (specific density of smoke) was evaluated using Eq.(1): where V is the incineration chamber volume (m 3 ), A is the specimen burning area (m 2 ), L is the light path length though the smoke (m), T is the light transmittance (%), and F is the temperature correlation coefficient.

Evaluation of Emission Amount
The emission factor of particulate matter (EF p ) was calculated using Eq. ( 2): ( / ) where Q is the flue gas flow rate (m 3 /min), S is the mass of dust collected on the filter (g), Q p is the flow rate of dust collection (m 3 /min), and M is the total combustible mass (kg).PM 10 and PM 2.5 emissions were evaluated first and PAH and heavy metals in the total particulate matter were calculated by dividing the emission mass by the consumed sample mass.The results are expressed as milligrams of emission per kilogram of aliquot [mg pollutant/kg waste].
The emissions were estimated based on the activity rate and the emission factors of each substance.

Management of Air Quality, Seoul Metro
In 2005, the Ministry of Environment established the Bureau of Metropolitan Atmosphere to improve air quality in metropolitan regions, including Seoul.While this active management strategy has improved air quality, more than 57.2% of Koreans currently reside in this region.Therefore, real environmental problems, including air pollution as well as a number of social concerns, exist at all times.There are various sources of air pollution, and open burning is an important local emission source.Individual municipalities and counties, pursuant to national regulations, prohibit all types of open burning.Nevertheless, local municipalities may not adopt stringent rules prohibiting all types of uncontrolled burning, but instead regulate open burning by addressing specific problems regarding public health.Although a few rules restrict incinerating arbitrary wastes, unsanctioned open burning is common due to ambiguous clauses or regulations.
Differing from U.S. law, Korean law describes illegal uncontrolled incineration of waste rather than providing a list of permissible materials and burning sites.In addition to burning agricultural residue, undetermined materials are consistently incinerated in metropolitan areas.Two critical reasons for this burning seem to be stringent recycling policies and meter-rate waste collection.Most open burning, therefore, has become illegal.Scientists have estimated that open burning, including all types of abnormal incineration, contributes up to 31.1% of PM 2.5 to metropolitan areas, based on receptor modeling (Lee et al., 2006).

Frequency of Open Burning
The field reality and type of combustibles were determined through a comprehensive public survey.We used the information obtained to estimate open burning activity and evaluate pollutant emissions.This survey was conducted based on the residential population by classifying the area into urban and rural districts.Survey questions included the frequency of open burning, the type of materials, volume and location of the burning, and the primary reasons for incineration.A total of 804 residents in urban and rural areas near the Seoul metro area and 38 public officers in charge of environmental affairs in local government were interviewed.The results indicate that more than 10.6% of households in the urban area and 45.2% in the rural area conduct open burning.
The most frequently burned wastes were wastepaper and vinyl, waste lumber and plastics, rubber and leather debris, and other food residues.These results are quite similar to the results of surveys conducted between 1998 and 2006.More than 80% of residents burn less than 10 L at a time and 10.8% of residents burn approximately 10 L. Open burning occurred 2 to 3 times a month in 67.1% of households surveyed, and 7.7% of households burn waste more than twice a week.The burning time was less than 1 hour for 82% of households surveyed.These results match the opinions of environmental officials working in the survey regions.Therefore, we chose sample waste and burning conditions based on the survey results.

Specific Density of Smoke (Ds)
Smoke is the particulate matter released from burning any combustible.The evolution of smoke, including prolonged burning time and maximum intensity, varied with the type of material burned (Fig. 3).The specific density of smoke (Ds) was relatively high for rubber and styrofoam, while the Ds of rice hulls was the lowest.Synthetic resins, including plastic and styrofoam, ignited easily and quickly reached the maximum Ds.Rice hulls, a typical agricultural residue, did not have a high Ds value, but consistently released smoke during combustion.The time required to reach maximum Ds was 13 minutes for paper, 3 minutes for rubber, and 16 minutes for fabric.Plastics and styrofoam approached the maximum Ds in a short time, 10 minutes, but within 30 minutes the amount of smoke generated decreased to less than 50% of the maximum.These results indicate that large amounts of fine particulates can be emitted in a short combustion time.Repeated tests of fabrics indicated a two-stage combustion, which may consist of the initial combustion materials and other materials burned during a subsequent combustion.Therefore, smoke is generated in two steps, each with a relatively low intensity.
In conclusion, the specific smoke density (Ds) followed the order of styrofoam > rubber > plastics > wood > fabrics > paper > agricultural residue for all combustion samples.The specific smoke densities provide emission estimates of particulate matter, including smoke from open burnings, and the combustion pattern can also be estimated based on the burning conditions.

Emission Factors
Since uncontrolled combustion of MSW in an open yard is carried out under various conditions, the emission factors of each pollutant may not provide adequate estimates of the average emissions.Nevertheless, we used a systematic program to attempt to better understand the contribution of open burning to regional air pollution and to enable the development of a management policy.The emission factors of fine dust, including PM 10 , PM 2.5 , heavy metals, and PAHs, were quantified as the amount of harmful materials released per unit mass of combustibles.The average emission factors based on mass are summarized in Table 2.

Particulate Matter
The lab-scale experiments were carried out to study the emission characteristics of particulate matter using a stack sampler.A laser diffraction size analyzer (Malvern Mastersizer) was used to verify the particle classification of the stack sampler, which ensured an accuracy level of 98% for PM 10 and 92% for PM 2.5 .
The emission factors for TPM and PM 10 calculated using Eq.(2) were highest for waste plastics and relatively low for wood combustion, though the values for PM 2.5 shown in Table 2 are relatively similar.TPM in the flue gas from waste paper was composed of 72% PM 10 and 46% PM 2.5 .Waste wood released a TPM of 1000 mg/kg including 83% PM 10 and 50% PM 2.5 .Waste plastics generated 1700 mg of TPM with 88% PM 10 and 29% PM 2.5 .TPM from the MSW burning was composed of 80% PM 10 and 52% PM 2.5 .In other words, PM 2.5 was high in burning wastes, which probably increases the risk to human health.PM 2.5 occupied more than 60% out of PM 10 except waste plastics.Plastics comprising polymers discharged a large amount of particulate matter, including smoke, due to an extended period of incomplete combustion (Kim et al., 2009).
MSW, a complex of domestic wastes, produced a fine dust made up of 65% PM 2.5 among the PM 10 .The test aliquots were intrinsic since they were procured from a local waste depository; therefore, various residues and debris remained on the surface.These contaminants might, in part, influence the properties of the particulate matter in the flue gas.Most plastics were PET bottles, which have a surface coating or color printing.These coatings and printings affect heavy metal and PAH formation as well as fine dust from combustion.Waste wood is often stained with paint or other pigments and organic adhesives.Waste paper also greatly varies in the primary material and is often laminated with vinyl or printed in color with artificial designs.These additives and impurities that remain in waste might form

Heavy Metals
Heavy metal emissions were evaluated in terms of the constituents of the total particulate matter (TPM).The analysis included lead (Pb), Nickel (Ni), Copper (Cu), Cadmium (Cd), Chromium (Cr), and Zinc (Zn).The average total heavy metals (THM) are summarized in Table 2. Table 3 presents the high and low limits of each individual element.The respective THM emission factors of the test combustibles were highest in plastics, followed by MSW, paper, and wood, similar to PM 10 .Although the absolute amount is small, the heavy metal content in suspended particulate dust is significant (Yoo et al., 2002).Therefore, strictly controlling combustion smoke can reduce the presence of heavy metals in the atmosphere.
The most common heavy metal in waste plastics was Zn followed by Cr, Pb, Ni, Cu, and Cd.The high zinc content in flue dust from waste plastics is consistent with that seen in other studies (U.S. EPA, 1997;Yoo et al., 2002).The EF of zinc, however, varied greatly in the test incineration of raw samples, with the exception of the MSW for which EF remained between 13.29 and 14.16 mg/kg.Plastic combustion might produce high lead and chromium content depending on the sample plastics.The consistent Ni and Cr content in waste paper attracts particular attention since it is frequently burned.Variability of Ni and Cr content among various paper samples is lower than other heavy metals.

Polyaromatic Hydrocarbons (PAHs)
PAHs that include more than 2 benzene rings are primarily generated in the atmosphere from incomplete combustion.PAHs are classified into gas or solid phases after emission according to molecular weight and temperature.While PAHs with 2 to 4 aromatic rings are in the gas phase or adsorb onto solid surfaces, those with more than five rings are on the surface of solid particles that are adsorbed.Many PAHs are also defined as hazardous air pollutants (HAPs) by the U.S. EPA.
This study focused on high molecular weight PAHs, such as benzo(a)anthracene, chrysene, benzo(b,k)fluoranthene, and benzo(a)pyrene as summarized in Table 4, due to their risk to environmental health (Lerda, 2010;NIOSH, 2010).TPM and PM 2.5 were collected on separate filters during test combustion, and the PAH content of each dust was determined.Table 2 indicates lower levels of PAHs on each gram of TPM than of PM 2.5 .This result was due to their specific surface area and adsorption capacity.Since the surface area to volume ratio of PM 2.5 is more than four times larger than TPM, the adsorption capacity of harmful PAHs is much higher.Burning of waste plastics released a large amount of PAHs.
MSW incineration releases a large amount of various PAHs, including particulate phenanthrene, fluoranthene, and pyrene (Table 4).Naphthalene and acenaphthylene, gaseous PAHs, were also detected.Naphthalene, which is highly sublimated, attracts particular attention.Carcinogenic benzo(a)pyrene appeared in some amount in every sample, but more significantly in waste plastics.
The emissions from MSW in this study were more than three times the total PAHs emissions reported by Kakareka (2003).Despite the different composition of household wastes, PAH constituents were quite similar.Both wastes produced a large amount of phenanthrene and did not release a detectable amount of dibenz(a,j)anthracene.Significant amounts of acenaphthylene, phenanthrene, pyrene, and benzo(a)pyrene were also found in both studies.

Amount of Emissions
Source tests or continuous emission monitors can determine the actual pollutant contribution from specific sources, such as open burning.The results of these simulated tests, however, are only applicable for estimating emissions if a reasonable activity rate of open burning can be determined.
Since determining an accurate activity to evaluate PM 10 , PM 2.5 , heavy metals, and PAHs was not easy, we used a public survey, national statistics, and foreign literature.First, the number of houses and residents as determined by the National Statistical Office was used to estimate the open burning activity based on the results of the survey in the methods sections.The results indicated that 10.6% of houses in urban areas and 45.2% of houses in rural areas periodically perform open burning.Therefore, the number  5. The amount of MSW that is burned was measured by the survey, which included questions about individual burning volume at any given instance.The survey showed that 11.7% of households have burned more than 10 L of waste at a time.From this comprehensive evaluation, the quantity of MSW burned was estimated to be 284 million L/yr.The apparent density of waste was used to obtain weight-based values.The density of domestic MSW was estimated by averaging monthly wastes for an 18-month period between 2008 and 2009, from Y-city, a city with both urban and rural areas in the Kyunggi province.The MSW aliquots of potential combustibles contained 23.2% paper, 30.9% plastics and vinyl, and 11.9% wood as well as incombustibles (Table 1).Our study found an average apparent specific gravity of 0.24.Therefore, the yearly MSW was 68 thousand tons in the Seoul metropolitan area of Korea.
An official environmental yearbook published by the Ministry of Environment presents the quantity of MSW released and the treatment method.These data enabled us to predict open burning activities, including the type of incineration.Moreover, these data can be used to evaluate the total emissions of harmful elements (MOE, 2006).Eight million tons of MSW were discharged in the metropolitan area in 1 year (basis: 2005), including 49.2% recycled materials.Therefore, 4.29 million tons of MSW are treated with official or unofficial processes every year.If 24.1% of this waste was disposed of by illegal or uncontrolled burning in open yards, according to the residential survey, approximately one million tons of waste disappears without any controlled treatment.The statistical data for waste generation and treatment were obtained from annual reports by local government and field industries according to waste management legislation.These data, however, are sometimes unreliable due to several problems, including insincere declarations and omission of illegal disposal.Private collectors underestimated the recycling rate, while waste treatment plants at landfills and incineration sites accumulate accurate data.In addition, illegal trash dumping as well as landfill and open burning that frequently occur in rural areas are not reflected in the official statistical figures.
The European Union (EU) provided average open burning activities for boundary countries, and many research institutes in the USA have studied these activities as well (U.S. EPA, 1997).Based on published literature, the EU and USA determined open burning rates to be between 0.25% and 0.3% of generated MSW near large cities.We estimated emissions based on 0.3%.Although the residential patterns Comparing these three methods, results based on the public survey and statistical data were significantly different.This difference could be due to the difficulties in accurately estimating the amount of MSW that was illegally treated, because current national statistics only contain officially treated wastes.As seen in Table 6, the emissions of harmful substances calculated from national data are much higher than the other two empirical methods.Therefore, this study used local survey results and data from other countries.
The emissions of harmful elements were evaluated using standard analyses as described above.Total emissions were evaluated based on the total quantity of waste burned in open yards in a metropolitan area near Seoul and the estimated amount of each type of emission is presented in Table 6.
Waste quality depends on lifestyle and topographical background, which vary between countries, so foreign models are not applicable.Therefore, surveying regional residents and national officers is the most reliable current method.
The amount of PM 2.5 discharged, which was evaluated by simple emission factors from MSW, was 46.6 tons every year in the metropolitan area, which is much less than predicted using national statistics.In other words, an uncertain amount of wastes from an official investigation are arbitrarily dumped and buried.The National Statistical Office collected environmental data for total suspended particulates (TSP), PM 10 , CO, and NO 2 .Atmospheric heavy metals, PAHs, and PM 2.5 have not been officially investigated by the government.According to national information, 15,533 tons of PM 10 were discharged in 2007 (MOE, 2008).Therefore, open burning contributes 0.44% of total PM 10 emissions in the metropolitan air pollution.Total heavy metal emissions were not significant enough to be of concern.Total PAHs discharged to the air by open burning were estimated to be 67 kg/yr, excluding agricultural debris.

CONCLUSIONS
To examine the effects of arbitrary open burning on air quality around the Seoul metropolitan area, the emission factors and quantities of major harmful substances were closely evaluated by public survey.Twenty-four percent of households in test areas were estimated to have customarily burned various wastes, such as paper, wood, plastic, and combustible MSWs in barrels or open yards near residential areas.The results demonstrated that artificial additives and contaminants on solid surfaces from manufacturing or disposal sources contain significant amounts of heavy metals.MSW samples were collected directly from a local depository and were given similar properties to simulate field materials disposed of by open burning.
Combustion characteristics were examined using a test combustion chamber; samples had different characteristics depending on the aliquot.Waste plastics produced the highest quantity of harmful substances of all the samples.The highest quantity of PAHs in MSW was in PM 2.5 ; one gram of PM 2.5 from MSW combustion produces 1.91 mg of PAHs.A total of 1,200 mg/kg of PM 10 and 780 mg/kg of PM 2.5 resulted from open burning of MSW.A total of 13.39 mg of heavy metals were released from burning 1 kg of MSW.Since defining the activity rate of open burning in a metropolitan area was not easy, three methods were used to estimate individual emissions.Public survey was the best predictor of burning activities.This study indicated that 71 tons of PM 10 are released every year, which accounts for 0.44% of metropolitan atmospheric dust, and 46.6 tons of PM 2.5 were produced from arbitrary open burning.Also, 914 kg/year of total heavy metals and 67 kg/year of PAHs were produced.

Fig. 2 .
Fig. 2. A schematic of the lab test system.
charge of environmental affairs in the local government was carried out to evaluate the amount of waste and types of open burning in residential areas.Test waste was selected according to the survey results and burned in a laboratoryscale incineration chamber.The emission factors of major harmful pollutants were then evaluated.The amount of emissions was estimated based on results from the survey and the literature.

Table 2 .
Summary of emission characteristics of hazardous substances.

Table 4 .
PAH mass content in TPM from MSW burning.

Table 5 .
Waste volume burned in open yards each year based on survey.

Table 6 .
Yearly emissions of harmful air pollutants from MSW open burning (tons/year).Direct survey, (II) National statistics, (III) Foreign statistics *PAH is based on TPM and reasons for illegal burning differ from Korea, this rate was used for comparison and resulted in approximately 40.6 tons of TSP per year, which creates 0.23% of the atmospheric dust in the metro area.