Assessment of VOC Emissions from Municipal Solid Waste Composting

Introduction

In 2008 the number of people living in urban areas exceeded the number of people living in rural areas; it is estimated that, in 2050, 70% of the population will live in cities inhabited by more than 10 million people (National Geographic, 2011). The accumulation of population around urban zones forces us to manage large amounts of municipal solid waste (MSW). Landfills have been the preferred option in many countries, but the available space for the construction of dumps is limited, and in the middle to long term, other solutions must be proposed. Incineration has been one of the management strategies considered and used for a long time, but current waste-management policies that establish hierarchies that prioritize reuse and recycling before disposal solutions (European Parliament, 2008), coupled with issues related to atmospheric contamination, arise serious doubts about the future viability of these processes.

Organic matter accounts for 50–60 wt% of the MSW, and therefore this fraction is liable for exploitation (Saeed et al., 2009; Turan et al., 2009). Composting is an attractive, feasible solution because it enables the acquisition of a profitable product, which serves as an organic amendment and can improve soil conditions. Atmospheric contamination and the amount of waste are decreased, and the biological cycle is preserved (Pagans et al., 2006).

To reach good conditions in the composting process, the facilities need an adequate design that considers the direction and intake of air, water supplies, location, environmental conditions, and other aspects related to the functioning and maintenance of the system. These factors are crucial for the rapid development of the process through aerobic degradation by transforming the organic matter and reducing the production of volatile organic compounds (VOCs) (D'Imporzano et al., 2008; Maulini-Duran et al., 2013), which mainly occur in the early stages of the process. Deficient or incomplete ventilation of the composting piles generates anaerobic conditions under which odorous sulfide compounds can be produced (Muezzinoglu, 2003; Zhang et al., 2013). Under aerobic conditions, the predominant volatiles are alcohols, ketones, esters, organic acids, and hydrocarbons, particularly toluene and ethylbenzene (Komilis et al., 2004; Pagans et al., 2006; Rodriguez-Navas et al., 2012). Dimethyl sulfide, dimethyl disulfide, limonene, and a-pinene are the major compounds that have been identified as emissions during the composting process (Scaglia et al., 2011; Rodriguez-Navas et al., 2012), although emissions may vary according to the type of residue, the state of decomposition, and the operating conditions of the process (Bruno et al., 2007; Shen et al., 2012). The production and emission of odorous or dangerous compounds is one of the greatest disadvantages of composting (Pagans et al., 2007).

Toxicological relevance of VOCs is a result of their high volatility and solubility; VOCs enter the organism through inhalation and absorption through the skin and affect primarily the central and peripheral nervous systems (Vallejo and Baena, 2007). The adverse effects on health that most VOCs can produce at high concentrations have been documented. VOC exposure has been proven to cause the following effects: (1) in the short term, topical reactions, such as acute irritation of the eyes, nose, throat, and skin; headache and difficulty focusing; dizziness; visual disorders; fatigue; loss of coordination; nausea; and memory disorders (Blount et al., 2006), and (2) in the long term, carcinogenicity (Ott et al., 2006) and damage to the liver, kidneys, and central nervous system. The toxicity of odorous hydrocarbides derived from benzene has been widely studied, demonstrating that benzene causes leukemia; toluene, ethyl benzene, and xylene cause peripheral neuropathies; and hydrocarbides in general cause liver and kidney damage (Vallejo and Baena, 2007).

It is generally accepted that VOC emissions, like bioaerosols and pathogenic microorganisms, involve a hazard for composting facility workers; thus, the exposure level is controlled to prevent potential negative health effects (Cadena et al., 2009; Nadal et al., 2009; Maulini-Duran et al., 2013). Composting is an automated process in the developed countries and monitoring is carried out by detectors and suitable equipment in the composting units, thereby the measurements indicating the actual exposure levels undergone by the workers. Nevertheless, operation is totally manual in the developing countries. Thus, the workers are placed on the upper level of the bed and turn it. Therefore, they are exposed to concentrations close to those corresponding to the equilibrium with air, which are much higher than those monitored by the measurement devices because the gaseous stream leaving the bed is rapidly diluted with the surrounding air prior to reaching the device.

However, in the literature, there are very few studies that have quantified the VOC emissions of composting facilities (Guo et al., 2004; Mao et al., 2006; Nadal et al., 2009; Delgado-Rodriguez et al., 2011; Kumar et al., 2011; Ryu et al., 2011; Scaglia et al., 2011; Rodriguez-Navas et al., 2012; Maulini-Duran et al., 2013), and in every case, the measurements have been made by gas-sample analysis. The selection of the gas sampling point is not straightforward, given that it should be indicative of the exposure level undergone by the workers (the upper surface of the pile in the compost facilities located in the developing countries), the choice and availability of instruments (vacuum pumps, absorption layers, etc.), the complexity of the instruments, and conservation of samples for transport and analysis (Li et al., 2013). In addition, sampling results can be influenced by both atmospheric conditions and the fast dissolution of contaminants leaving the pile.

Therefore, the aim of this work is to propose and test a complementary methodology for determining the VOC concentration in the air inside a compost pile, given that this value provides a suitable indication of the risks undergone by the personnel in the manually operated compost plants when turning over the piles. This methodology is based on taking solid samples (easier to process and send to laboratories without being altered) and quantifying the contaminants in the solid phase (compost) by desorption. Further, a calculation method based on the equilibrium between phases is proposed for determining the VOC concentration in the gaseous phase (air), and the risk associated with the exposure of these vapors is quantified.

Materials and Methods

The composting process was carried out in a facility located in the village of Altavista, municipality of Medellín, which produces and markets the compost. The facility is located at 1,600 m above sea level and has an average ambient temperature of 22°C throughout the whole year. One thousand three hundred kilograms of organic waste separated at the source by the inhabitants was used, and pruned plant material (39 kg) was added to reduce the moisture content. The MSW used basically consists of fruit and vegetable wastes that are well classified by the consumers. It is therefore a rather clean material, with a considerable particle size and irregular texture, which makes up beds of high bed voidage. The bulk density is measured weekly and increases from 80 to 180 kg/m³, which ensures suitable bed voidage values throughout the whole process.

The compost was set in an array of three-by-five beds with dimensions of 1.50×1.50 m², which are filled to a height between 1 and 1.5 m. The waste was manually turned and the bedding was changed twice per week to ensure adequate ventilation. The process had a total duration of 7 consecutive weeks. During turning, the compost was weighed to determine mass loss, and after the new bed was formed, the height was recorded.

Samples of ∼500 g were taken weekly at nine different points of the pile. The samples were stored in sealed bags and sent to the laboratory for analysis. Before sampling, the temperature was recorded at each point. To follow and confirm the composting process, the following physicochemical analyses were performed: moisture content, particle density and bulk density, water-holding capacity, ash content, electrical conductivity, pH, total nitrogen content, total organic carbon content, cationic exchange capacity, and levels of phosphorus and metals. Moreover, the samples were subjected to microbiological characterization (for enterobacteria, molds, and yeasts and mesophilic and thermophilic organisms), and mineralization of the solid phase was measured by respirometry, following standard procedures in each case. All the analyses carried out confirmed the proper development of the composting process. The evolution of temperature indicates the existence of two thermophilic phases: a first severe one, with a peak at 80°C in the second week, and a second one with temperature being slightly lower than 60°C in the fifth week. The moisture content decreases steadily during the composting process to values close to 40 wt%, and bulk density almost doubles, reaching a value of 160 kg/m³ at the end of the process. Finally total nitrogen content increases slightly along the process to a value of 1.05 wt% in the final product.

Identification and quantification of VOCs were carried out on two types of samples: (1) gas delivered by the bed and collected in 5-L Tedlar sampling bags, according to the procedure of Wilkins and Larsen (1996), which consisted in placing a cylindrical tube in the central part of the bed that is connected to the bag, which is removed after 24 h, and (2) solid-phase microextraction (SPME) of the compost sample according to the methodology proposed by Peñalver-Hernando (1999). For this purpose, 0.5 g of solid sample (compost) was introduced into a 40-mL vial (clear vial, screw top hole cap with PTFE/silicone septa) that is placed on top of a suitable extraction fiber (SPME 75 μm Carboxen PDMS for Merlin Microseal, 23-gauge needle Manual Holder Black; Sigma-Aldrich, Saint Louis, MO), and the mixture was heated on a plate at 60°C for 30 min. The temperature chosen corresponds to an average value in the pile for the composting process. Further, preliminary assays showed that 30 min is sufficient for reaching equilibrium at this temperature. The holder was then removed and the extracted products were analyzed. All the analyses have been carried out in triplicate in order to verify reproducibility of results.

The VOCs in both samples were identified with a gas chromatograph Agilent 7890A GC with a PTV injector coupled to a mass spectrometer MS Agilent5975C VL-MSD triple axis. A capillary column DB-624 (30 m×1.40 μm×0.25 mm) was used. A 10 μL of volume from the Tedlar bag samples was injected, and for the SPME, the fiber was left to desorb for ∼5 min. The initial column temperature was 40°C for 5 min, and then the sample was heated with a ramp of 9°C/min to 175°C and then a ramp of 45°C/min to 200°C and held at this temperature for 3 min. The injector and detector were maintained at 160°C and 230°C, respectively.

Results and Discussion

Gases contained in the Tedlar bags as well as those desorbed by SPME have been weekly analyzed. As an example, Fig. 1 shows the chromatograms obtained from the analysis of samples that correspond to (a) the Tedlar bag and (b) SPME; both obtained on the eighth day of the composting process. One can observe the formation of volatile compounds, which demonstrates that the SPME enables detection of a greater number of volatile compounds compared to the Tedlar bags, in a ratio of 40:1.

OPEN IN VIEWER

FIG. 1.

Chromatograms of (a) Tedlar bag sample and (b) solid-phase microextraction sample, both obtained on the eighth day of the composting process.

For the GC/MS analysis of samples obtained by SPME, compounds were detected for each of the 7 weeks of the composting process. The largest numbers of volatile compounds were seen in the first weeks of the process, 67 compounds in week 1 and 158 compounds in week 4 (Table 1), confirming that the largest formation of volatile compounds occurs in the first weeks of the process (Komilis et al., 2004; Shen et al., 2012). Overall, an increase in the production of volatile compounds was detected until the fourth week and then detected a gradual decrease from week 5 until the end of the process, in whose sample was detected a total of 63 compounds.

Aromatic compounds, such as toluene, xylenes, ethylbenzene, naphthalene derivatives, styrene, benzaldehyde, and indole, were recorded in greater quantities during the third and fourth weeks. Aliphatic hydrocarbons, such as propanoic acid derivatives, and terpenes, such as limonene, pinene, and camphor, were identified during the entire composting process but especially in the first and second weeks, which is consistent with the results obtained in previous studies of the same type of compost waste (Wilkins and Larsen, 1996; Komilis et al., 2004; Schlegelmilch et al., 2005; Nadal et al., 2009).

Health risk assessment requires the concentrations of compounds harmful to health in the gas phase, that is, gaseous emissions from the bed of compost, because they are the ones that affect the workers through inhalation. However, the results obtained by SPME provide the amount of a compound in the compost, not in the surrounding air. Gas-liquid-solid equilibrium should be taken into account for estimating gas concentrations. In fact hydrocarbons are partitioned between the solid and liquid, so that \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*}M_{VOC} = M_{VOC}^s + M_{VOC}^l \tag{1}\end{align*} \end{document}

Where M_VOC is the total hydrocarbon mass, M^s_VOC is the VOC mass in the solid phase, and M^l_VOC is the VOC mass on the liquid phase.

M^s_VOC and M^l_VOC can be determined by Equations (2) and (3). \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*}M_{VOC}^s = SM_{comp} \tag{2}\end{align*} \end{document}

Where S is the VOC concentration in the solid phase (in mg VOC/kg compost) and M_comp is the compost mass (bed mass). \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*}M_{VOC}^l = \O V_{com} \ L \tag{3}\end{align*} \end{document}

Where Ø is the volume humidity that is calculated in the humid base, V_com is the compost volume (solid phase) of the bed, and L is the VOC concentration in the liquid phase (in mg VOC/m³ liquid phase).

Ø is obtained from the weight humidity (H) expressed as the humid base and is measured from the weekly compost sample characterization trials. \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*}\O = \frac { HM_ { comp } } { \rho_w V_ { bed } ( 1 - \varepsilon ) } \tag { 4 } \end{align*} \end{document}

Where ρ _w is water density, V_bed is the bed volume, and ɛ is the porosity, which is determined from the real compost density (ρ) and the density of the bed (ρ _b ). \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*}\varepsilon = 1 - \frac { \rho_b } { \rho } \tag { 5 } \end{align*} \end{document}

In Equation (3), V_com=V_bed (1–ɛ), and L=S/K_DT, where K_DT is the VOC distribution constant between liquid and solid phases. Although a rigorous application of the method proposed in this article requires further studies in order to determine the value of the partitioning constant for each of the VOCs detected in a complex matrix, such as the compost, this article assumes a typical value of 0.01 m³/kg for monoaromatic hydrocarbons (Eweis et al., 1998).

Combining Equations (1) through (5), we obtain \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*}S = \frac { M_ { VOC } } { M_ { comp } \left( \frac { H } { \rho_w K_ { DT } } + 1 \right) } \tag { 6 } \end{align*} \end{document}

As shown, Equation (6) enables to determine the concentration of VOCs in the solid phase exclusively from the results of the SPME-GC/MS tests, the mass of the bed, and the humidity of the bed. From the values of S, the values of L are obtained, and with these, the concentration of VOCs in the gas phase (G, mg VOC/m³ air), which is the concentration of hydrocarbons in the air that is inhaled by the worker when he turns the bed. \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*}G = \frac { LM } { k_H^o \exp \left( \frac { dlnk_H } { d ( 1 / T ) } \left( \frac { 1 } { T } - \frac { 1 } { T^o } \right) \right) P_ { atm } \ RT } \tag { 7 } \end{align*} \end{document}

In Equation (7), M is the molecular mass of the hydrocarbon, k_H^o is Henry's constant at the standard temperature T^o (298.15 K), dlnk_H/d(1/T) is the temperature dependence constant (K), T is the temperature of the bed (K), P_atm is the atmospheric pressure (in bar), and R is the universal gas constant.

Although the methodology proposed in the previous paragraphs can be applied to each of the compounds detected, a statistical analysis (SPSS 15.0) was performed to establish which compounds are the most significant in the health risk assessment. Based on the results of the statistical analysis and the demonstrated toxicity of aromatic hydrocarbons, xylenes, ethylbenzene, and toluene were selected for the health risk assessment using the new quantification methodology proposed before. These compounds have not been classified as carcinogens by the International Agency for Research on Cancer (IARC), but they have recognized toxic effects, so their presence in emissions from the bed can be considered potentially dangerous, particularly for workers of the manually operated composting facility who are in direct contact with the air inside the bed when it is turned.

The assessment requires precisely and accurately determining the total amount of these compounds present in the solid phase of the bed (compost) from the analysis of the samples. Accordingly, a standard mixture of 125 ppm o-, m-, and p-xylene, ethylbenzene, and toluene was prepared and subjected to the same treatment as that for the sample analysis. Quantification of these aromatic compounds in the compost samples was performed by the external pattern method. Finally the mass of xylene, ethylbenzene, and toluene in the bed at the different weeks is obtained by integration, based on the results obtained by SPME-GC/MS in the nine sampling bags extracted from the different points in the pile. The results are shown in Table 2.

Results in Table 2 correspond to the total mass in the compost bed for each one of the VOCs studied. Applying the methodology proposed in this article and described before, the concentration of these compounds in the air inside the bed can be estimated. Table 3 shows the results for the samples weekly analyzed.

In the literature there are no data obtained by directly measuring the VOC concentration in the manually operated composting facilities, and therefore the results obtained in this study are compared with others in studies carried out in MSW plants and automated composting plants. In the analysis of urban waste disposal bins, Statheropoulos et al. (2005) measured the concentration of these and other compounds and found that the maximum value obtained in autumn (average temperature 28.3°C and 47.4% relative humidity) from four overloaded waste bins that were left uncollected for 7 days was 7.8 μg/m³ for m-xylene, 12.7 μg/m³ for ethylbenzene, and 8.1 μg/m³ for toluene. The values are much lower than those in this work, but in addition to differences in the sampling method, waste disposal bin conditions are not comparable to those of the composting bed. Durmusoglu et al. (2010) analyzed benzene, xylenes, ethylbenzene, and toluene over 10 sampling points from a landfill facility and found maximum values of 5.67 mg/m³ for xylene, 3.72 mg/m³ for ethylbenzene, and 10.23 mg/m³ for toluene; these results are similar to those obtained in this study, although the mean values for the three pollutants were 0.34, 0.24, and 1.27 mg/m³, respectively. These authors conducted a comprehensive literature review of similar studies of landfill gas (Eklund et al., 1998; Schweigkofler and Niessner, 1999; Kim et al., 2006, 2008), and the values are similar to those obtained in this work, even though the sampling method and the substrate (gas) are different. Mao et al. (2006) studied the generation of odors in the three largest food waste composting plants in Taiwan but focused their study on the ambient air inside the plants, exhaust outlets, and plant boundaries, so that the maximum levels reported are well below those in this study: 81 μg/m³ of xylenes, and 29–64 μg/m³ for ethylbenzene and toluene, respectively. In a more recent study, Nadal et al. (2009) and Scaglia et al. (2011) measured the presence of microbiological agents and chemical pollutants in indoor air samples collected from an MSW organic fraction in a treatment plant. Most of the sampling and analysis of VOCs were carried out according to the NIOSH 1501 guide, and the highest levels of xylenes, ethylbenzene, and toluene were obtained in sorting bins with values of 2.85, 0.78, and 4.77 mg/m³, respectively. These values are very similar to those obtained in this work and help validating the methodology developed and presented in this article.

After calculating the concentration of xylenes, ethylbenzene, and toluene, air in equilibrium with the compost will be used as a standard to quantify the potentially adverse health effects of noncarcinogenic compounds (Durmusoglu et al., 2010). The toxicity assessment is made through the hazard ratio (HR), defined as the ratio of daily intake (I) and the reference dose (RfD), so that values >1 reflect the danger of an adverse health effect. I is estimated with Equation (8). \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*}I = \frac { G*IR* \cdot EF*ED } { BW*AT } \tag { 8 } \end{align*} \end{document}

In Equation (8), G is the contaminant concentration (mg/m³), IR is the inhalation rate (m³/day), EF is the exposure frequency (day/year), ED is the exposure duration (year), BW is the body weight (kg), and AT is the averaging time (day). The US EPA (1989) recommends standard values of 70 kg and 20 m³/day for BW and IR, respectively. To estimate EF and ED and considering that the risk is calculated for composting facility workers, the mean labor duration can be established as 40 h/week. Considering annual vacation (4 weeks), EF for workers is 80 days [=40×(52–4)/24]. In addition, the work lifetime, and therefore ED, is generally assumed to be 35 years. As contaminant exposure is averaged over a lifetime, the AT value of 70 years (or 25,500 days) can be assumed. RfD represents the level of daily intake of a particular substance that should not produce an adverse health effect (LaGrega et al., 1994) and can be deduced from the reference concentration values (RfC) available on the Internet from the Integrated Risk Information System website (US EPA). In this work, reference values of 0.029 mg/(kg·day⁻¹) for xylenes, 0.29 mg/(kg·day⁻¹) for ethylbenzene, and 1.43 mg/(kg·day⁻¹) for toluene were used. Table 4 shows the RH values that were calculated for the three monoaromatic hydrocarbons for each week of the composting process.

Results suggest that the production of xylenes, ethylbenzene, and toluene during the process can be considered a risk for workers, which makes necessary to take appropriate occupational health measures to prevent the health problems that these contaminants could cause for workers in the facilities. Due to the high concentrations, the predominant route of exposure is usually by inhalation and effects may occur in less time and may be more acute. On the other side, this work has quantified the production of three aromatic hydrocarbons and confirmed the presence of many VOCs in considerable concentration levels. Synergistic effects cannot be ruled out and they may cause more damage than each compound could cause separately. In addition, the consequences of exposure must be considered cumulative over time in the absence of information indicating otherwise (Durmusoglu et al., 2010).

As concluded by Nadal et al. (2009), these results confirm that occupational exposure to VOCs is a health risk in composting facilities; therefore, protective equipments, at least integrated filter masks and gloves, are highly recommended for composting facility workers.

Conclusion

VOC emissions in a manually operated composting facility have been determined by GC/MS analysis, based on two types of samples: gas delivered by the bed and retained in a 5-L Tedlar bag and also estimated by means of SPME of solid samples. The second one is more appropriate for detecting a great number of volatile compounds than the Tedlar bag sampling. Thus, the former is able to detect an amount 40 times greater than the latter. Further, other advantages of the second methodology are the simplicity and the ease of storage and transport of the samples, and the reliability, robustness, and level of standardization of analysis techniques. The results obtained allow estimating the amount of VOCs in the pile by applying solid-liquid-gas phase equilibrium calculations. The results obtained following this procedure have been compared with those in other studies on the emission of VOCs from MSW or composting facilities, and they prove the validity of the methodology proposed. Further the results confirm health risk of exposure to xylenes in manually operated MSW composting facilities.