Particle-Scale Modeling of Methane Emission ... - ACS Publications

Article pubs.acs.org/est

Particle-Scale Modeling of Methane Emission during Pig Manure/ Wheat Straw Aerobic Composting Jinyi Ge, Guangqun Huang, Jing Huang, Jianfei Zeng, and Lujia Han* Laboratory of Biomass & Bioprocessing Engineering, College of Engineering, China Agricultural University, (East Campus), Box 191, Beijing 100083, China S Supporting Information *

ABSTRACT: Inefficient aerobic composting techniques significantly contribute to the atmospheric methane (CH4) levels. Macro-scale models assuming completely aerobic conditions cannot be used to analyze CH4 generation in strictly anaerobic environments. This study presents a particle-scale model for aerobic pig manure/wheat straw composting that incorporates CH4 generation and oxidation kinetics. Parameter estimation revealed that pig manure is characterized by high CH4 yield coefficient (0.6414 mol CH4 mol−1 Cman) and maximum CH4 oxidation rate (0.0205 mol CH4 kg−1 VSaero h−1). The model accurately predicted CH4 emissions (R2 = 0.94, RMSE = 2888 ppmv, peak time deviation = 0 h), particularly in the self-heating and cooling phases. During mesophilic and thermophilic stages, a rapid increase of CH4 generation (0.0130 mol CH4 kg−1 VS h−1) and methanotroph inactivation were simulated, implying that additional measures should be performed during these phases to mitigate CH4 emissions. Furthermore, CH4 oxidation efficiency was related to oxygen permeation through the composting particles. Reducing the ambient temperature and extending the aeration duration can decrease CH4 emission, but the threshold temperature is required to trigger the self-heating phase. These findings provide insights into CH4 emission during composting and may inform responsible strategies to counteract climate change.

■

INTRODUCTION According to the latest World Meteorological Organization (WMO) Greenhouse Gas Bulletin, because of increased emissions from anthropogenic sources, the globally averaged mole fraction of methane (CH4) reached a new high of 1824 ± 2 ppb in 2013.1 Manure management contributes approximately 7.9−9% of global anthropogenic CH4 emission to the atmosphere;2 CH4 emission during manure-based composting could account for 2−3% of the total carbon content of the composting materials and has thus caused concern. 3,4 Mathematical modeling and simulation could be practical tools for improving our understanding of the mechanism of CH4 emission and for prediction of CH4 evolution during aerobic composting, thereby providing theoretical guidance for reducing the environmental impact of CH4. Macro-scale models assuming completely aerobic conditions cannot be used to analyze CH4 generated from a strictly anaerobic environment.5 Hamelers6,7 presented a particle-scale aerobic composting model. A schematic can be found in our previous work8 to visualize the aerobic composting mechanism at the particle scale. Gaseous oxygen (O2) is converted to dissolved O2 and diffuses into the composting particles through the gas/liquid interface. Heterotrophs use O2 and soluble substrates for metabolism, forming an aerobic layer where aerobic reactions are predominant. Along the diffusion path, © XXXX American Chemical Society

consumption of O2 by aerobic microorganisms results in reduction of the O2 concentration to an extremely low level; this causes the particle to have an anaerobic core in which the insoluble substrate can be hydrolyzed to soluble products. The anaerobic core in the composting particle could explain the generation of anaerobic gases during aerobic composting, and the assumption of the anaerobic/aerobic coprocess provides a basis for investigating the mechanism of CH4 emission.9,10 Ge et al.11 justified the existence of the aerobic layer and anaerobic core of manure-based composting particles using Fourier transform infrared microspectroscopy (FTIRM) and characterized the variations in the thickness of the aerobic layer (Lp). The size and shape of the manure particles were also investigated.12 On the basis of the above particle characteristics, Ge et al.8 improved the particle-scale oxygen uptake rate (OUR) model for the aerobic composting of pig manure and wheat straw. These findings provide a methodological and theoretical foundation for the development of a particle-scale CH4 emission model for manure-based aerobic composting. Received: August 26, 2015 Revised: February 11, 2016 Accepted: April 5, 2016

A

DOI: 10.1021/acs.est.5b04141 Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Article

Environmental Science & Technology Jäckel et al.13 and Lou and Nair14 suggested that CH4 could be oxidized to carbon dioxide (CO2) by aerobic microorganisms during its diffusion from the anaerobic core. Hence, the mechanism of CH4 emission for aerobic composting includes CH4 generation in the anaerobic core of the composting particle and CH4 oxidation in the aerobic layer. However, in terms of aerobic composting processes, macroscale CH4 emission models predominate and use experimental data to obtain empirical emission factors for composting heaps under different conditions.15 To our best knowledge, no studies of the mechanisms of CH 4 emissions during aerobic composting have been published. To explore the kinetics of CH4 generation, some studies on anaerobic degradation can be used for reference as follows.16−18 CH4 generation during anaerobic digestion is a sequential process involving hydrolysis, acidogenesis, acetogenesis, and methanogenesis, with hydrolysis as the rate-limiting step.16 A study investigating landfilling of municipal solid wastes found that methane generation rate (vgen) is correlated with the hydrolysis rate (RSi).17 After examining anaerobic digestion of biowastes, Veeken and Hamelers18 suggested that CH4 production may be proportional to the hydrolysis rate of particulate organic matter when no intermediary products are accumulated. For the kinetics of CH4 oxidation, studies of biofiltration19,20 and landfilling21−23 suggested that the methane oxidation rate (voxi) follows Michaelis−Menten kinetics and can be dramatically influenced by the composting temperature (T). Moreover, voxi has also been shown to be closely related to the OUR of heterotrophic bacteria that degrade organic matter. This relationship can be explained by the competition between methanotrophs and heterotrophs.19,20 The aim of this study was to develop a particle-scale CH4 emission model that specifies the kinetics of CH4 generation and oxidation for the aerobic composting of pig manure and wheat straw. The interaction between CH4 emission and operation parameters was examined by sensitivity analysis. This effort is expected to facilitate the design and optimization of composting strategies that comply with tight environmental regulations.

represents the initial volatile solids, and t is the composting time (h). In the aerobic layer of the particle, voxi followed Michaelis− Menten kinetics,19−23 voxi =

(3)

vmax CH4,out K m + CH4,out

β (4)

Correction of Kinetic Parameters. The model quantified the influence of both temperature and OUR on key kinetic parameters, including kh(T), vmax, and Km. First, Van Lier et al.30 and Ge et al.31 demonstrated that high temperature would contribute to CH4 generation, probably because kh(T) was proportional to T, which could be depicted by the Arrhenius equation:6,7 k h(T ) = k h(Tr) e−E / R(1/ T + 273 − 1/ Tr + 273)

(5)

where Tr is the reference temperature (°C), kh(Tr) is the hydrolysis rate constant at the reference temperature (h−1), E is the activation energy (kJ mol−1), and R is the ideal gas constant (kJ mol−1 °C1−). For the kinetics of CH4 oxidation, the temperature correction factors for the maximum methane oxidation rate (f_max_T) and the half saturation constant (f_Km_T) were described by eq 621,22 and eq 7,23 respectively. ⎧0 0°C < T ≤ 5°C ⎪ 5°C < T ≤ 15°C ⎪ 0.0142T ⎪ ⎨ 15°C < T ≤ 33°C f _vmax _T = 0.112T − 1.47 ⎪ ⎪ 2.235 − 0.18(T − 33) 33°C < T ≤ 45°C ⎪ ⎩0 T > 45°C (6)

f _K m _T = (0.00678 + 0.009814T ) e−1700(1/ T + 273.15 − 1/298.15)

(7)

Besides the temperature, the activity of heterotrophs indicated by the OUR value was also shown to influence the CH4 oxidation kinetics owing to the competition for O2 and substrate between heterotrophs and methanotrophs. The OUR correction factor for the maximum methane oxidation rate ( f_vmax_OUR) could be characterized as follows,19,20

(1)

f _vmax _OUR = 1.2835 e−44.1698OUR

6,28

R Si = k h(T )Si,0 e−kht

β

vemit = YCH4k h(T )Si,0 e−kht −

MODEL DEVELOPMENT Model Assumptions. Pig manure is considered the major reactant while straw, because of relatively slow biodegradation, acts as a filler to maintain and ensure proper porosity of the composting mixture.24−26 Furthermore, the penetration and distribution of O2 in composting mixtures are adequate and relatively homogeneous. Our previous work27 showed that Lp reflects the particle porosity parameter crucial for gaseous emissions and aerobic conditions. Kinetics of CH4 Emission. Considering the correlation between vgen and RSi during anaerobic degradation,17,18 methane yield coefficient (YCH4) is proposed to quantify CH4 production in the anaerobic core of the composting particle: RSi was driven by first-order reaction kinetics,

K m + CH4,out

where vmax is the maximum methane oxidation rate (mol CH4 kg−1 VSaero h−1), VSaero represents the volatile solids in the aerobic layer of the particle, Km is the half saturation constant (ppmv), and β is the mass conversion factor of VS to VSaero (kg VSaero kg−1 VS). On the basis of the kinetics of CH4 generation and oxidation as described by eqs 1−3, the methane emission rate (vemit) was the difference between vgen and voxi,13,29

■

vgen = YCH4R Si

vmax CH4,out

(8)

Notably, the applicable range is 0 < OUR < 0.0500 mol O2 kg−1 VS h−1.20 Energy Balance Equation. The simulated values of T feeding into eqs 5−7 could be obtained by solving the energy balance equations for the aeration and nonaeration phases. The energy balance equations can be found in our previous work.8

(2) −1

where kh(T) is the hydrolysis rate constant (h ), Si,0 is the initial insoluble substrate concentration (mol Cman kg−1 VS), Cman represents the carbon content of pig manure, VS B


Article

Environmental Science & Technology

Figure 1. Schematic of the aerobic composting system: (1) composting reactor, (2) mixing system, (3) material inlet, (4) sampling opening, (5) material outlet, (6) water supplement, (7) gas outlet, (8) off-gas absorption, (9) built-in temperature sensor, (10) built-in oxygen sensor, (11) gas inlet, (12) air and water distribution system, and (13) automatic control system.

Particle-Scale OUR Model. The simulated values of OUR feeding into eq 8 could be obtained with the help of the particle-scale OUR model:8 OUR =

∫L

Lmax

G (L , λ , γ )

s

+

∫0

Ls

where RCman is the rate of increase of pig manure carbon content (mol Cman kg−1 VS h−1), RSs is the rate of increase of soluble substrate (mol Cman kg−1 VS h−1), RXa is the rate of increase of aerobic microbe mass (mol C kg−1 VS h−1), and YSi is the biomass yield on insoluble substrate (mol C mol−1 Cman). More detailed calculations are specified in Table S1 in the Supporting Information.

Lp 1 R X dL L YO2 a

G (L , λ , γ )

1 YSsR Si dL YO2

■

(9)

EXPERIMENTAL SECTION Aerobic Composting Experiment. Temperature or O2 controls and more advanced methods based on online measurements of respiration (OUR) are used in forced-aerated composting practice36,37 to reduce energy consumption and relieve CH4 emissions by improving O2 penetration and distribution in composting mixtures. However, in this study, we chose an on/off aeration mode system because of the size limitation of the lab-scale reactor used for the composing experiments (100 L) described and the requirement for CH4 and OUR measurements to be executed at the end of an aeration phase. Preliminary experiments verified an adequate and sufficiently homogeneous O2 supply during this aeration mode. The optimized aeration strategy was 1 h “on” and 1 h “off”, with 0.02 m3 kg−1 VS h−1 aeration rate. Pig manure was collected from the livestock and poultry test site of the Chinese Academy of Agricultural Sciences (Changping, Beijing, China). Wheat straw was collected from suburban areas of Beijing and chopped into 3−5 cm lengths for use as a bulking agent. Three aerobic composting experiments (experiments A, B, and C) were performed; each experiment had a duration of 21 days, utilized a ventilation pattern of 1 h on/1 h off, and had an aeration rate (Q) of 0.02 m3 kg−1 VS h−1. Data obtained from experiments A and B were used for parameter estimation of YCH4 and vmax, and data from experiment C were employed for model validation. The mass ratios of raw materials applied in the three experiments differed; however, other operation conditions were kept as identical as possible. For experiment A, 49.0 kg of pig manure and 7.0 kg of wheat straw were mixed thoroughly. The mixture was loaded into a

where Lmax is the maximum particle radius (m), Ls is the switch radius (m), G(L, λ, γ) is the probability density function of the gamma distribution (dimensionless), L is the particle radius (m), λ is the scale parameter (m−1), γ is the shape parameter (dimensionless), YO2 is the biomass yield of oxygen (mol C mol−1 O2), RXa is the growth rate of aerobic microbes (mol C kg−1 VS h−1), and YSs is the biomass yield of soluble substrate (mol C mol−1 Cman). Further details regarding these model parameters are described in the Table S1 of the Supporting Information. Transformation and Degradation of Organic Matter. The organic matter of composting particles comprises three substrates, i.e., soluble, insoluble, and inert, consistent with the studies of Oudart et al.32,33 Soluble substrate consists of easily biodegradable carbohydrates, some simple amino acids, and low-molecular-weight acids.24,25,34 Insoluble substrate is less degradable and contains hemicellulose-like matter. X-ray diffraction (XRD) determined that lignin-like structures comprise the most inert fraction of pig manure.27 Detailed definitions and calculations of these three fractions are provided in Table S1 in the Supporting Information. Organic matter transformations mainly concern hydrolysis of the insoluble substrate, generation and microbial consumption of the soluble substrate, and incorporation of dead microbes into the insoluble substrate. These processes have been successfully simulated in pig slurry, wheat straw and pig manure composting experiments.32,33 Therefore, quantification of the pig manure degradation is now described by the equation:35 R Cman = −R Si + R Ss +

1 RX YSi a

(10) C


Article

Environmental Science & Technology Table 1. Physicochemical Characteristics of Composting Materials composting mixture index MCa

OMb

CT/NTb

FASb

OURb

experiment A B C A B C A B C A B C A B C

pig manure

wheat straw

± ± ± ± ± ± ± ± ±

6.67 ± 0.42 5.58 ± 0.51 7.68 ± 0.57 93.10 ± 0.23 92.28 ± 0.35 93.06 ± 0.55 90.00 ± 8.96 88.94 ± 2.16 94.97 ± 4.56

71.69 68.10 65.34 82.65 72.65 80.70 13.13 14.93 14.50

0.86 1.10 1.53 0.63 0.58 0.78 0.19 1.71 1.20

day 0 62.66 ± 61.62 ± 60.99 ± 85.90 ± 76.93 ± 83.27 ± 18.40 ± 17.63 ± 19.87 ± 47.14 49.07 47.76 0.0049 0.0019 0.0040

day 3

1.10 1.60 1.88 0.80 0.37 0.80 0.73 0.73 0.50

65.23 ± 66.00 ± 69.23 ± 81.30 ± 71.30 ± 80.00 ± 15.80 ± 17.43 ± 17.15 ± 48.20 50.00 48.47 0.0908 0.1037 0.0895

day 6

2.60 3.98 3.20 0.80 0.98 0.94 0.44 0.44 0.43

51.23 ± 60.34 ± 56.23 ± 74.30 ± 64.30 ± 78.80 ± 11.89 ± 13.58 ± 13.36 ± 49.53 50.85 48.71 0.0433 0.0570 0.0466

2.90 2.97 2.88 0.60 0.87 1.12 0.23 0.43 0.33

day 9

day 15

day 21

46.80 ± 3.20 56.80 ± 3.22 50.30 ± 3.20 72.00 ± 0.77 62.00 ± 1.11 77.00 ± 0.98 9.91 ± 0.27 11.23 ± 0.42 12.46 ± 0.56 49.89 51.06 49.06 0.0269 0.0342 0.0271

44.00 ± 2.30 51.00 ± 2.80 49.00 ± 2.90 72.60 ± 0.52 62.60 ± 0.76 77.00 ± 1.12 9.16 ± 0.17 10.27 ± 0.54 13.31 ± 0.59 49.80 51.01 49.06 0.0129 0.0235 0.0178

44.00 ± 2.00 48.00 ± 2.90 47.00 ± 3.30 72.00 ± 0.63 62.00 ± 0.68 76.00 ± 1.60 9.26 ± 0.19 10.40 ± 0.53 14.66 ± 0.53 49.89 51.06 49.24 0.0195 0.0218 0.0197

a Measurement based on wet weight. bMeasurement based on dry weight. MC is the moisture content (%), OM is the organic matter content (%), CT is the total carbon content (%), NT is the total nitrogen content (%), FAS is the free air space (%), and OUR is the oxygen uptake rate (mol O2 kg−1 VS h−1).

100-L cylindrical reactor (2.00 height × 0.45 internal diameter m; Figure 1).38 Samples of the raw pig manure, wheat straw, and initial mixtures were collected and frozen at −4 °C until analysis. For experiment B, the masses of pig manure and wheat straw were 50.0 and 5.0 kg, respectively. For experiment C, the masses of pig manure and wheat straw were 49.0 and 5.0 kg, respectively. Additionally, 3.0 kg of deionized water was added to the materials obtain a moisture content of 50−67%.39 Chemical Analysis and Measurement. The moisture content (MC) and organic matter content (OM) of the raw materials and initial mixtures in the three experiments were measured according to standard procedures.40 The CT/NT ratio was calculated from the total carbon (CT) and total nitrogen (NT), both of which were determined by dry combustion using an elemental analyzer (Vario MACRO; Elementar, Germany). Chemical properties were determined using three replicates. Free air space (FAS) was employed to assess the porosity of the composting mixture and can be calculated as39,41 FAS = 1 −

m _mix(100% − MC_mix) V _mixρw Gs

1 OM_mix 100% − OM_mix = + Gs Gv Gf

intervals for gas measurement and analysis were determined as follows: every 6 h between 5 and 95 h and every 16 h between 95 and 495 h. The O2 concentration (volume fraction) of the input and output streams was measured at the end of the aeration phase using an O2 sensor (HD5; Nanjing Huideng Electronic Technology Co., Ltd., China), the response time and measurement accuracy of which was less than 1 min and ±1.5%, respectively. Simultaneously, gaseous samples of the input and output were collected in 1-L gas bags. The CH4 concentration in the bags was immediately analyzed by gas chromatography using a flame ionization detector (FID; GC2014; Shimadzu, Japan) and an RT-Q-BOND column (30 length m × 0.53 diameter mm). Operation conditions were set as follows: column temperature, 50 °C; injector temperature, 120 °C; and FID temperature, 200 °C. The carrier gas was nitrogen with a flow rate of 5 mL min−1 at a pressure of 31.2 kPa. Data Processing and Calculation. The experimental OUR and vemit could be calculated based on differences in the gas concentrations (O2 and CH4) between the input and output streams.43−45

(11)

OUR = 1000Q (O2,in − O2,out )/Vm

(13)

vemit = 1000Q (CH4,out − CH4,in)/Vm

(14)

(12)

where m_mix is the initial mass of composting materials (kg), MC_mix is the initial moisture content of composting materials (%), V_mix is the composting material volume (m3), and ρw is the water density (kg m−3). Gs, Gv, and Gf (dimensionless) are specific gravities of the mixture, volatile fraction, and fixed fraction, respectively. OM_mix is the initial organic matter content of composting materials (%). The composting temperature (T) and ambient temperature (Ta) were automatically monitored at 20 min intervals using Pt100 thermocouples located in the middle of the composting materials and the surrounding atmosphere, respectively. Data from the thermocouples were documented using a programmed data acquisition system (DT85; DataTaker Pty Co., Ltd., Australia). Given the possible significant variation in the emissions of O2 and CH4 during the early stage of composting,39,42 time

where 1000 is the conversion factor from m3 to L, O2,in is the oxygen concentration in the input stream (%), O2,out is the oxygen concentration in the output stream (%), Vm is the molar volume at 1 atm at 25 °C (L mol−1), CH4,in is the methane concentration in the input stream (ppmv), and CH4,out is the methane concentration in the output stream (ppmv). A, B, and C are three independent experiments, with slightly different manure to straw mixing ratios. Experiments A and B were used to estimate YCH4 and vmax. Data from experiment C were employed for model validation. Parameter estimation was performed with the nlinfit function in Matlab software (Mathworks, Inc., MA). The optimum result was achieved when the deviations between real data and simulations were minimized. D


Article

Environmental Science & Technology The CH4 emission model was written using Matlab software, and the differential equations were numerically solved using the Runge−Kutta algorithm-based ODE45 routine in this package. The quality of the fit and the deviation between simulated and experimental values were evaluated by the determination coefficient (R2) and root-mean-square error (RMSE), respectively. The relevant calculations can be found in our previous work.8 Sensitivity Analysis. The kinetic parameters (YCH4 and vmax) were modified to ±25% and ±50% of their default values. The influence of these two parameters on the maximum CH4,out, the cumulative CH4 emission, and the proportion of oxidized CH4 was investigated. The proportion of oxidized CH4 was calculated as the ratio of the amount of oxidized CH4 to the total CH4 generation. With regard to the operation parameters, the values of Ta and Q were modified to ±25% and ±50% of their default values, while aeration duration (on) and nonaeration duration (off) were modified to 0.25, 0.50, 6.00, and 12.00 h. The effects of these four parameters on the cumulative CH4 emission, the proportion of oxidized CH4, the duration for which the composting temperature was higher than 50 °C (Length50), and the time at which the OUR reached 0.0125 mol O2 kg−1 VS h−1 (Time0.0125) were then investigated. Length50 describes the duration of thermophilic phase, and at Time0.0125, the composting materials were nearly matured.46 For example, the influence of changing one parameter to −25% on Length50, denoted as ΔLength50,−25, was quantified as follows: ΔLength50, −25 = (Length50, −25 − Length50,0)/Length50,0 × 100%

(15)

Figure 2. (a) Ambient temperatures (Ta) and composting temperatures (T) in experiments A, B, and C. (b) Oxygen and methane concentration in the output stream (O2,out and CH4,out, respectively) in experiments A, B, and C.

where Length50,0 is the simulated value of Length50 with the default value of the parameter (h), and Length50,−25 is the simulated value of Length50 with a modified value of the same parameter (h).

■

first 2 weeks contributed about 95% to the total amount. Hao et al.3 and Maeda et al.4 demonstrated that the total CH4 emission from composting cattle manure accounted for 2−3% of the carbon content in the manure; in contrast, the value in this study was around 4%, implying that pig manure had higher potential to generate CH4.25 Parameter Estimation. Simulations of the CH4,out values for experiments A and B are illustrated in Figure 3 and Table 2. The modeled results fit well with the experimental data, particularly in the self-heating and cooling phases. The mean estimated values of YCH4 and vmax of pig manure based on the experimental data were 0.6414 mol CH4 mol−1 Cman and 0.0205 mol CH4 kg−1 VSaero h−1, respectively. The estimation of YCH4 is close to the previously reported value (0.54 mol CH4 mol−1 Cman).25 However, both our current results and the result from the previous study are much higher than that reported for brewery-spent grains;16 this is because manure contains more carbohydrates, proteins, and lipids, which feature higher YCH4 values.48 The vmax value of pig manure is higher than that of landfill cover soil (0.0044 mol CH4 kg−1 VS h−1),49 probably because this parameter is proportional to the organic matter content within a certain range.21,50 Model Validation. The simulations of T, OUR, and CH4,out for experiment C are illustrated in Figure 4 and Table 3. Overall, the model provided a good prediction of the variation

RESULTS AND DISCUSSION Chemical Analysis and Aerobic Composting. The evolutions of all physicochemical characteristics are illustrated in Table 1. FAS values exceeded 45% for the three composting mixtures. This was maintained for 21 days, which supports the model assumption that the penetration and distribution of O2 in composting mixtures are adequate and relatively homogeneous. The evolutions of chemical properties, especially for CT/ NT, between experiments A and B are not completely comparable. Therefore, experiments A and B are now treated as two independent experiments in the manuscript to allow YCH4 and vmax estimations. We averaged the estimates from the two experiments to reduce the impact of the variability of pig manure chemical properties. In the experimental data for Ta, T, O2,out, and CH4,out during pig manure/wheat straw aerobic composting (Figure 2), T proceeded through mesophilic, thermophilic, cooling, and maturation phases. O2,out was always greater than 10%, maintaining an aerobic environment. CH4,out changed dramatically between day 0 and day 9, exhibiting an immediate increase followed by a rapid decrease. This variation is consistent with the composting of sludge47 and dairy manure.42 Moreover, CH4 emission mainly occurred during the first half of the composting period. Numerically, CH4 emission over the E


Article


Owing to these changes in CH4 generation and oxidation, CH4 emission mainly occurred during the first half of the experimental period. Thus, additional mitigation measures should be performed during this stage. As shown in Figure 4c, CH4,out peaked on day 3, which could be precisely simulated. However, simulations of CH4,out during the cooling phase yielded higher values than the experimental data, possibly because YCH4 was overestimated, as it ignored the flow of carbon from the hydrolysis step to other products, for example, volatile fatty acids (VFAs), and losses from physical processes, such as diffusion. As the composting materials stabilized and matured, CH4,out remained at a low level, which was well predicted. The lower CH4 production and emission is probably due to the digestible organic matter being almost exhausted in the thermophilic phase.32 Although vgen in the cooling phase (15 and 25 °C) was low, the cumulative emissions from this phase occupied almost 51% of the total CH4 production. Comparison between RCman simulations and the experimental carbon content of composting mixture is given in Figure 4d. These two parameters differ since the carbon content of the composting mixture also incorporates carbon originating in the wheat straw. On the other hand, similar trends between pig manure and composting mixture carbon contents indicated that the degradation of carbon in the composting mixture was mainly attributed to the decomposition of pig manure. It helps to justify the model assumption that pig manure is the major reactant. After quick decomposition of the soluble substrate, the carbon content decreased during the mesophilic phase.55 Following this phase, numerous microbes were inactivated due to the high temperature and were converted to the insoluble substrate, which led to increased carbon content. The soluble substrate was almost exhausted while the hydrolysis of insoluble substrate proceeded through the cooling stage. For the composting of pig manure and wheat straw, the maximum voxi and the proportion of oxidized CH4 were 0.93 × 10−3 mol CH4 kg−1 VS h−1 and 10.34%, respectively. Jäckel et al.13 calculated these two values by incubating the biowaste samples under ambient air in 120 mL serum bottles, yielding measurements of 4.1 × 10−3 mol CH4 kg−1 dw h−1 (dw represents dry matter) and 46%, respectively. Both of these values are higher than the results obtained in our study, probably because the incubation process permitted greater access to O2 than typical composting practices, which could lead to larger Lp values in waste particles and higher oxidation efficiency. Sensitivity Analysis. The effects of YCH4 and vmax on the maximum CH4,out, the cumulative CH4 emission, and the proportion of oxidized CH4 are presented in Table 4. Higher YCH4 values indicated greater CH4 production capacity. Thus, the maximum CH4,out and cumulative CH4 emission increased almost proportionally with the increase in YCH4, and the

Figure 3. Simulation of methane concentration in the output stream (CH4,out) for (a) experiment A and (b) experiment B.

in the three variables and the peak time. The deviation in the peak time of T (16 h) was at the bottom of the previously reported range (0.1−4.5 days),51 although the temperature of the thermophilic phase was overestimated. The mean deviation in OUR was 0.0130 mol O2 kg−1 VS h−1, which is much lower than the previously reported values (0.0275−0.0558 mol O2 kg−1 VS h−1).51 However, the model could not simulate the gradual decrease in OUR caused by changes in different microbial species.52 The results of the simulations correspond to the welldocumented phenomenon that CH4 generation is normally faster than its oxidation. As shown in Figure 4b, the simulated vgen increased to 0.0130 mol CH4 kg−1 VS h−1 with increasing temperature, in agreement with the notion that most methanogens are mesophilic and thermophilic.13,53,54 In contrast, the simulated voxi during the initial period was close to zero because the temperature optimum of methanotrophic bacteria (33 °C)21,22 has been exceeded. A cover of finished compost, which has lower temperature and can exert a biofiltering effect, can decrease extensive CH4 emissions during the thermophilic phase.13

Table 2. Estimates of the Methane Yield Coefficient (YCH4) and Maximum Methane Oxidation Rate (vmax) and Simulation Accuracy for Pig Manure/Wheat Straw Aerobic Composting (Experiments A and B)a experiment

YCH4 (mol CH4 mol−1 Cman)

vmax (mol CH4 kg−1 VSaero h−1)

RCH42

RMSECH4 (ppmv)

A B

0.5711 0.7117

0.0108 0.0302

0.74 0.82

1140 1118

a

Cman represents the carbon content of pig manure. VSaero represents the volatile solids in the aerobic layer of the particle. RCH42 and RMSECH4 are the determination coefficient and root-mean-square error of methane concentration in the output stream. F


Article


Figure 4. Simulations of (a) composting temperature (T) and oxygen uptake rate (OUR), (b) rates of methane (CH4) generation and CH4 oxidation, and (c) methane concentration in the output stream (CH4,out). (d) Experimentally determined carbon content of the composting mixture and simulations of the pig manure carbon content.

were improved and reduced, respectively. These data suggested that the CH4 emission model was sensitive to YCH4 and vmax. Therefore, it is necessary to characterize these two kinetic parameters for each type of composting material. Sensitivity analyses of the operation parameters on the cumulative CH4 emission, the proportion of oxidized CH4, Length50, and Time0.0125 are listed in Table 5. High Ta values induced greater CH4 emission because the growth of methanogens was accelerated. This finding was consistent with observations of composting for cattle manure,56 sludge,47 and household wastes.57 In contrast, a much lower Ta cannot trigger and maintain the self-heating phase very well, leading to a smaller Length50. Given the trade-off between CH4 emission and inertness of microorganisms, Ta could be optimized to 27.5 °C (i.e., +25%). Excessive aeration could dramatically reduce CH4 emission but also caused increasing heat loss, consistent with the composting of manure and barley straw.58 For the laboratory-scale composting system used in this study, the optimal Q value was 8.34 L min−1 (i.e., +25%). Appropriately

Table 3. Simulation Accuracy for Pig Manure/Wheat Straw Aerobic Composting (Experiment C)a evaluation index

R2

RMSE

deviation of peak time (h)

T OUR CH4,out

0.89 0.83 0.94

8.8 °C 0.0191 mol O2 kg−1 VS h−1 2888 ppmv

16 −6 0

a 2

R is the determination coefficient, RMSE is the root-mean-square error, T is the composting temperature, OUR is the oxygen uptake rate, and CH4,out is the methane concentration in the output stream. VS represents the initial volatile solids. The simulated peak time was subtracted by the experimental value to obtain the deviation.

proportion of oxidized CH4 decreased accordingly. Changing vmax barely affected the maximum CH4,out because when CH4,out peaked, the methanotrophs were already inactivated by the high temperature. Higher vmax values suggested that the sample had enhanced potential for oxidation of CH4. As a result, the proportion of oxidized CH4 and the cumulative CH4 emission

Table 4. Sensitivity Analysis of the Methane Yield Coefficient (YCH4) and Maximum Methane Oxidation Rate (vmax)a YCH4

a

vmax

relative change (%)

−50

−25

+25

+50

−50

−25

+25

+50

maximum CH4,out cumulative CH4 emission proportion of oxidized CH4

−50.00 −51.23 21.32

−24.99 −25.85 9.96

25.01 26.26 −8.67

50.00 52.80 −16.19

0.00 5.19 −44.96

0.00 2.46 −21.34

0.00 −2.21 19.18

0.00 −4.23 36.70

CH4 is methane, and CH4,out is the methane concentration in the output stream. G


Article

Environmental Science & Technology Table 5. Sensitivity Analysis of Operation Parametersa ambient temperature −25

aeration rate

relative change (%)

−50

+50

−50

cumulative CH4 emission proportion of oxidized CH4 Length50 Time0.0125

−14.88 19.30 −15.87 0.00

−7.54 8.70 9.45 −2.79 −6.35 6.35 0.00 1.29 aeration duration

22.20 −28.27 14.29 1.29

126.62 −18.71 22.22 5.16

relative change (%)

0.25 h

0.50 h

6.00 h

12.00 h

cumulative CH4 emission proportion of oxidized CH4 Length50 Time0.0125

23.28 −50.24 −72.12 5.28

20.02 −39.68 −55.52 −9.00

−19.74 22.79 44.36 5.21

−22.36 22.31 52.38 5.21

+25

−25

+25

+50

24.40 −24.94 −8.28 10.71 1.59 −7.94 1.29 0.00 nonaeration duration

−39.34 9.97 −9.52 0.00

0.25 h

0.50 h

6.00 h

12.00 h

−12.38 20.20 21.36 5.86

−8.23 11.09 −8.88 −9.00

25.10 −66.32 −43.19 26.34

9.77 −72.44 −65.33 46.39

a Length50 is the duration for which the composting temperature was higher than 50 °C, and Time0.0125 is the time at which OUR reached 0.0125 mol O2 kg−1 VS h−1. VS represents the initial volatile solids.

■

extending the duration of aeration could help to alleviate the environmental impact of composting. Long nonaeration duration (12.00 h) not only aggravated CH4 emission by reducing the proportion of oxidized CH4 by 72.44% but also increased the Time0.0125 value (the time at which the OUR reached 0.0125 mol O2 kg−1 VS h−1 and the composting materials were nearly matured46) by 46.39%. In this case, the stability and quality of the end product may be diminished when the composting was completed according to the original plan. Therefore, the optimal ventilation pattern was on 6 h/off 15 min. These findings could provide insights into the mechanism of CH4 emission during manure-based aerobic composting and theoretical guidance for designing environmentally friendly and economical operation strategies.

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.est.5b04141. Main parameters and variables of the methane model and references (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*Fax: 86-10-6273-6778. Phone: 86-10-6273-6313. E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

■

ENVIRONMENTAL RELEVANCE According to the model validation results, more attention should be focused on CH4 emission during the early stages of composting. Improving O2 permeation through the composting particles may enhance CH4 oxidation efficiency. Sensitivity analyses demonstrated that appropriately reducing ambient temperature and extending aeration duration could reduce CH4 emission. These findings improve our understanding of the mechanisms of CH4 emission during manure-based aerobic composting and may provide insights to facilitate decisionmaking regarding responsible strategies for mitigating climate change. The applicability of the model to other waste composting practices, for instance, composting of wastewater sludge, that are known to lead to production of significant CH4 emissions,59 should be addressed and verified in other studies. First, the inputted particle parameters of different composting materials, such as Lmean, Lmax, and Lp, may differ and have to be specifically characterized. Furthermore, O2 concentration gradients exist in large-scale compost heaps, toward the heap surface,33 and need to be considered. Indeed, O2 concentrations are higher in layers closer to the surface than in the heap center, and it has been shown that CH4 produced in the bottom layers might be oxidized during transport to the surface.13 Besides, inhibition effects resulting from ammonia build-up on CH4 emissions comprise a valid concern as well.60 Therefore, the potential competition between nitrifiers, denitrifiers, methanogens, and methanotrophs should be characterized in the future work to make the model closer to actual conditions.

ACKNOWLEDGMENTS This work was financially supported by the Program for Changjiang Scholars and Innovative Research Team in University (Project No. IRT1293), the National Key T ec h n o lo g y S up p o r t P r o g r a m o f Ch i n a ( G r a n t 2015BAC02B02-02), and the International Science & Technology Cooperation Program of China (Grant 2015DFA90370).

■

REFERENCES

(1) Greenhouse Gas Bulletin: The State of Greenhouse Gases in the Atmosphere Using Global Observations through 2013; World Meteorological Organization: Geneva, Switzerland, 2014; https://www.wmo. int/pages/mediacentre/press_releases/documents/1002_GHG_ Bulletin.pdf. (2) Ruirui, C.; Yiming, W.; Junli, H.; Xiangui, L. Methane emission and mitigation strategies in animal manure management system. Acta Pedologica Sinica 2012, 49 (4), 815−823. (3) Hao, X.; Chang, C.; Larney, F. J. Carbon, nitrogen balances and greenhouse gas emission during cattle feedlot manure composting. J. Environ. Qual. 2004, 33 (1), 37−44. (4) Maeda, K.; Hanajima, D.; Morioka, R.; Toyoda, S.; Yoshida, N.; Osada, T. Mitigation of greenhouse gas emission from the cattle manure composting process by use of a bulking agent. Soil Sci. Plant Nutr. 2013, 59 (1), 96−106. (5) Cayuela, M. L.; Sánchez-Monedero, M. A.; Roig, A.; Sinicco, T.; Mondini, C. Biochemical changes and GHG emissions during composting of lignocellulosic residues with different N-rich byproducts. Chemosphere 2012, 88 (2), 196−203. (6) Hamelers, H. V. M. Modeling composting kinetics: A review of approaches. Rev. Environ. Sci. Bio/Technol. 2004, 3 (4), 331−342.

H


Article

Environmental Science & Technology (7) Hamelers, H. V. M. A mathematical model for composting kinetics. Ph.D. Dissertation, Wageningen University, Wageningen, The Netherlands, 2001. (8) Ge, J.; Huang, G.; Huang, J.; Zeng, J.; Han, L. Modeling of oxygen uptake rate evolution in pig manure−wheat straw aerobic composting process. Chem. Eng. J. 2015, 276, 29−36. (9) D’Imporzano, G.; Crivelli, F.; Adani, F. Biological compost stability influences odor molecules production measured by electronic nose during food-waste high-rate composting. Sci. Total Environ. 2008, 402 (2−3), 278−284. (10) Maulini-Duran, C.; Puyuelo, B.; Artola, A.; Font, X.; Sánchez, A.; Gea, T. VOC emissions from the composting of the organic fraction of municipal solid waste using standard and advanced aeration strategies. J. Chem. Technol. Biotechnol. 2014, 89 (4), 579−586. (11) Ge, J.; Huang, G.; Yang, Z.; Huang, J.; Han, L. Characterization of the dynamic thickness of the aerobic layer during pig manure aerobic composting by Fourier transform infrared microspectroscopy. Environ. Sci. Technol. 2014, 48 (9), 5043−5050. (12) Ge, J.; Huang, G.; Huang, J.; Zeng, J.; Han, L. Quantitative characterization of the particle size and shape of pig manure. Trans. CSAM 2015, 46 (3), 192−196. (13) Jäckel, U.; Thummes, K.; Kämpfer, P. Thermophilic methane production and oxidation in compost. FEMS Microbiol. Ecol. 2005, 52 (2), 175−184. (14) Lou, X. F.; Nair, J. The impact of landfilling and composting on greenhouse gas emissions−a review. Bioresour. Technol. 2009, 100 (16), 3792−3798. (15) Brown, S. Greenhouse gas accounting for landfill diversion of food scraps and yard waste. Compost Sci. Util. 2016, 24 (1), 11−19. (16) Vavilin, V. A.; Fernandez, B.; Palatsi, J.; Flotats, X. Hydrolysis kinetics in anaerobic degradation of particulate organic material: an overview. Waste Manage. 2008, 28 (6), 939−951. (17) Lay, J.; Li, Y.; Noike, T. Mathematical model for methane production from landfill bioreactor. J. Environ. Eng. 1998, 124 (8), 730−736. (18) Veeken, A.; Hamelers, B. Effect of temperature on hydrolysis rates of selected biowaste components. Bioresour. Technol. 1999, 69 (3), 249−254. (19) Mancebo, U.; Hettiaratchi, J. P. A.; Hurtado, O. Compost as a growth medium of methanotrophic bacteria. J. Solid Waste Technol. Manag. 2010, 36 (3), 751−760. (20) Mancebo, U.; Hettiaratchi, J. P. A.; Hurtado, O. D. Study on the correlation between dissolved organic carbon, specific oxygen uptake rate, and exchangeable nitrogen and the performance of granular materials as support media for methanotrophic biofiltration. J. Hazard., Toxic Radioact. Waste 2014, 18 (1), 11−15. (21) Abichou, T.; Mahieu, K.; Chanton, J.; Romdhane, M.; Mansouri, I. Scaling methane oxidation: from laboratory incubation experiments to landfill cover field conditions. Waste Manage. 2011, 31 (5), 978− 986. (22) King, G. M.; Adamsen, A. P. S. Effects of temperature on methane consumption in a forest soil and in pure cultures of the methanotroph Methylomonas rubra. Appl. Environ. Microbiol. 1992, 58 (9), 2758−2763. (23) Sadasivam, B. Y.; Reddy, K. R. Landfill methane oxidation in soil and bio-based cover systems: a review. Rev. Environ. Sci. Bio/Technol. 2014, 13 (1), 79−107. (24) Tremier, A.; de Guardia, A.; Massiani, C.; Paul, E.; Martel, J. L. A respirometric method for characterising the organic composition and biodegradation kinetics and the temperature influence on the biodegradation kinetics, for a mixture of sludge and bulking agent to be co-composted. Bioresour. Technol. 2005, 96 (2), 169−180. (25) Møller, H. B.; Sommer, S. G.; Ahring, B. K. Methane productivity of manure, straw and solid fractions of manure. Biomass Bioenergy 2004, 26 (5), 485−495. (26) Li, Z.; Lu, H.; Ren, L.; He, L. Experimental and modeling approaches for food waste composting: a review. Chemosphere 2013, 93 (7), 1247−1257.

(27) Ge, J.; Huang, G.; Huang, J.; Zeng, J.; Han, L. Mechanism and kinetics of organic matter degradation based on particle structure variation during pig manure aerobic composting. J. Hazard. Mater. 2015, 292, 19−26. (28) Vasiliadou, I. A.; Muktadirul Bari Chowdhury, A. K.; Akratos, C. S.; Tekerlekopoulou, A. G.; Pavlou, S.; Vayenas, D. V. Mathematical modeling of olive mill waste composting process. Waste Manage. 2015, 43, 61−71. (29) Hantush, M. M. Modeling nitrogen−carbon cycling and oxygen consumption in bottom sediments. Adv. Water Resour. 2007, 30 (1), 59−79. (30) Van Lier, J. B.; Martin, J. L. S.; Lettinga, G. Effect of temperature on the anaerobic thermophilic conversion of volatile fatty acids by dispersed and granular sludge. Water Res. 1996, 30 (1), 199−207. (31) Ge, H.; Jensen, P. D.; Batstone, D. J. Relative kinetics of anaerobic digestion under thermophilic and mesophilic conditions. Water Sci. Technol. 2011, 64 (4), 848−853. (32) Oudart, D.; Paul, E.; Robin, P.; Paillat, J. M. Modeling organic matter stabilization during windrow composting of livestock effluents. Environ. Technol. 2012, 33 (19−21), 2235−2243. (33) Oudart, D.; Robin, P.; Paillat, J. M.; Paul, E. Modelling nitrogen and carbon interactions in composting of animal manure in naturally aerated piles. Waste Manage. 2015, 46, 588−598. (34) Vlyssides, A.; Mai, S.; Barampouti, E. M. An integrated mathematical model for co-composting of agricultural solid wastes with industrial wastewater. Bioresour. Technol. 2009, 100 (20), 4797− 4806. (35) Buendía, I. M.; Fernández, F. J.; Villaseñor, J.; Rodríguez, L. Biodegradability of meat industry wastes under anaerobic and aerobic conditions. Water Res. 2008, 42 (14), 3767−3774. (36) Scaglia, B.; Tambone, F.; Genevini, P. L.; Adani, F. Respiration index determination: dynamic and static approaches. Compost Sci. Util. 2000, 8 (2), 90−98. (37) Puyuelo, B.; Gea, T.; Sánchez, A. GHG emissions during the high-rate production of compost using standard and advanced aeration strategies. Chemosphere 2014, 109, 64−70. (38) Zhang, A.; Huang, G.; Zhang, S.; Yang, Z.; Han, L. Design and test on an experimental aerobic composting reactor system. Trans. CSAM 2014, 45 (7), 156−161. (39) Haug, R. T. The Practical Handbook of Compost Engineering; CRC Press: Boca Raton, FL, 1993. (40) Test Methods for the Examination of Composting and Compost; U.S. Composting Council: Bethesda, MD, 2002; http:// compostingcouncil.org/tmecc/. (41) Wang, Y.; Huang, G.; Han, L.; Zhang, A.; Ge, J. Effects of plastic and straw bulking agents on swine manure composting. Trans. CSAM 2013, 44 (5), 158−163. (42) Mulbry, W.; Ahn, H. Greenhouse gas emissions during composting of dairy manure: influence of the timing of pile mixing on total emissions. Biosyst. Eng. 2014, 126, 117−122. (43) Adani, F.; Lozzi, P.; Genevini, P. Determination of biological stability by oxygen uptake on municipal solid waste and derived products. Compost Sci. Util. 2001, 9 (2), 163−178. (44) Scaglia, B.; Erriquens, F.; Gigliotti, G.; Taccari, M.; Ciani, M.; Genevini, P.; Adani, F. Precision determination for the specific oxygen uptake rate (SOUR) method used for biological stability evaluation of compost and biostabilized products. Bioresour. Technol. 2007, 98 (3), 706−713. (45) Thompson, A.; Wagner-Riddle, C.; Fleming, R. Emissions of N2O and CH4 during the composting of liquid swine manure. Environ. Monit. Assess. 2004, 91 (1), 87−104. (46) Organic Soil Conditioners−Composts−Determination of Oxygen Uptake−Respirometric Method; BNQ: Sainte-Foy, Quebec, Canada, 2004; http://www.compost.org/pdf/ENGbnq.0413-200-5-A2.pdf. (47) Maulini-Duran, C.; Artola, A.; Font, X.; Sánchez, A. A systematic study of the gaseous emissions from biosolids composting: raw sludge versus anaerobically digested sludge. Bioresour. Technol. 2013, 147, 43−51. I


Article

Environmental Science & Technology (48) Ponsá, S.; Gea, T.; Sánchez, A. Anaerobic co-digestion of the organic fraction of municipal solid waste with several pure organic cosubstrates. Biosyst. Eng. 2011, 108 (4), 352−360. (49) Whalen, S. C.; Reeburgh, W. S.; Sandbeck, K. A. Rapid methane oxidation in a landfill cover soil. Appl. Environ. Microb. 1990, 56 (11), 3405−3411. (50) Wilshusen, J. H.; Hettiaratchi, J. P.; De Visscher, A.; Saint-Fort, R. Methane oxidation and formation of EPS in compost: effect of oxygen concentration. Environ. Pollut. 2004, 129 (2), 305−314. (51) Mason, I. G. Mathematical modelling of the composting process: a review. Waste Manage. 2006, 26 (1), 3−21. (52) Mohajer, A.; Trémier, A.; Barrington, S.; Martinez, J.; Teglia, C.; Carone, M. Microbial oxygen uptake in sludge as influenced by compost physical parameters. Waste Manage. 2009, 29 (8), 2257− 2264. (53) Vallero, M. V.; Camarero, E.; Lettinga, G.; Lens, P. N. Thermophilic (55−65 °C) and extreme thermophilic (70−80 °C) sulfate reduction in methanol and formate-fed UASB Reactors. Biotechnol. Prog. 2004, 20 (5), 1382−1392. (54) Su, X.; Wu, Y.; Xia, D.; Chen, X. Effect of temperature on biological methane generation of low rank coal. Coal Geol. Explorer 2012, 40 (5), 24−26. (55) Lin, Y. P.; Huang, G. H.; Lu, H. W.; He, L. Modeling of substrate degradation and oxygen consumption in waste composting processes. Waste Manage. 2008, 28 (8), 1375−1385. (56) Ahn, H. K.; Mulbry, W.; White, J. W.; Kondrad, S. L. Pile mixing increases greenhouse gas emissions during composting of dairy manure. Bioresour. Technol. 2011, 102 (3), 2904−2909. (57) Ermolaev, E.; Sundberg, C.; Pell, M.; Jönsson, H. Greenhouse gas emissions from home composting in practice. Bioresour. Technol. 2014, 151, 174−182. (58) Chowdhury, M. A.; de Neergaard, A.; Jensen, L. S. Potential of aeration flow rate and bio-char addition to reduce greenhouse gas and ammonia emissions during manure composting. Chemosphere 2014, 97, 16−25. (59) Åkerman, A.; Massagué, A.; Mehier, S.; Senante, E.; Escaler, I.; Martel, J. L.; Audic, J. M. Methodology for the assessment of the greenhouse gas impact of wastewater sludge treatment. J. Water Clim. Change 2010, 1 (4), 227−233. (60) Pagans, E.; Barrena, R.; Font, X.; Sánchez, A. Ammonia emissions from the composting of different organic wastes. Dependency on process temperature. Chemosphere 2006, 62 (9), 1534−1542.

J


Particle-Scale Modeling of Methane Emission ... - ACS Publications

Recommend Documents