Simultaneous removal of ammonia and volatile organic compounds from composting of dead pigs and manure using pilot-scale biofilter

ABSTRACT

Odor emission is one of the most common problems associated with dead animals composting. Biofiltration treatment for eliminating exhaust odors formed during dead pigs and manure composting has been studied. The composting and biofiltration process consisted of two series of tests. Composting experimental trials lasted 6 weeks, and composting was performed using six pilot-scale reactor vessels. A total of 37 kinds of volatile organic compounds (VOCs) present in the air were identified, and temporal variations were determined during the 42 days of composting. Dimethyl sulfide (DMS), dimethyl disulfide (DMDS), dimethyl trisulfide (DMTS), and trimethylamine (TMA) were identified as the main odors VOCs component according to odor active values (OAVs). Nine biofilter vessels containing mature compost were used in studying the effect of different (30, 60, and 100 s) empty bed retention times (EBRT) on the simultaneous removal efficiencies (REs) of NH₃, DMS, DMDS, DMTS, and TMA. Results indicated that the inlet concentration of NH₃ applied was 12–447 mg m⁻³, and the average removal efficiencies were 85.4%, 88.7%, and 89.0% for EBRTs of 30, 60, and 100 s, respectively. The average REs of DMS, DMDS, DMTS, and TMA were 79.2%–95.4%, 81.9%–94.0%, 76.7%–99.1%, and 92.9%–100%, respectively, and their maximum elimination capacity (ECs) were 220, 1301, 296, and 603 mg m⁻³ h⁻¹, respectively. The optimal EBRT for the stimulation removal of NH₃, DMS, DMDS, DMTS, and TMA was 60 s.

Implications: Dimethyl sulfide (DMS), dimethyl disulfide (DMDS), dimethyl trisulfide (DMTS), and trimethylamine (TMA) were identified as the main odors VOCs component during dead pigs and manure composting. Biofilter with mature as media can be used to stimulation remove NH₃, DMS, DMDS, DMTS, and TMA, the optimal empty bed retention times EBRT was 60 s.

Introduction

The disposal of dead animals is an important aspect of livestock farming (Berge et al. 2009; Murphy and Handwerker 1988), and the commonly used methods include burial, composting, rendering, and incineration. Composting has been recognized as an environmentally acceptable method (Ahn, Richard, and Glanville 2008; Henry 1995; Pagans, Font, and Sánchez 2005). Odor emissions and atmospheric pollution are the most common problems associated with composting. Ammonia (NH₃) is one of the main compounds responsible for generating offsite odors and pollution, especially when composting dead animals, which produces high amounts of nitrogen. The loss of nitrogen during composting normally implies the poor agronomical quality of the final compost and environmental pollution problems, because ammonia can cause vegetation stress, ecosystem alteration and has a role in formation of fine particulate matter (Buijsman, Maas, and Asman 1987; Krupa 2003; Shen et al. 2012). Intense microbial activity during composting may produce and release volatile organic compounds (VOCs) (Akdeniz et al. 2010; Smet, Van Langenhove, and De Bo 1999), which include a wide group of organic compounds, such as sulfur-containing and nitrogen-containing compounds, alcohols, phenols, ketones, esters, volatile fatty acids, and terpenes (Pagans, Font, and Sanchez 2006). Specifically, dimethyl sulfide (DMS), dimethyl disulfide (DMDS), dimethyl trisulfide (DMTS), and trimethylamine (TMA), are considered as main compounds found in VOCs emissions in composting facilities responsible of odor pollution (Hong et al. 2017; Leea et al. 2020; Park et al. 2013; Van Durme, McNamara, and McGinley 1992; Zhang et al. 2016a). NH₃ contributes the most to odor emission during aerobic composting (Zhu et al. 2016). The emission of DMS, DMDS, DMTS, and TMA is usually low, but these VOCs can cause a strong sense of odor due to their low olfactory threshold (Smet et al. 1996a). Therefore, mitigating the odor emissions is necessary to increase the environmental benefits of dead animals composting.

Biofilter is the most attractive current technology for odor treatment (Hong and Park 2004; Turan, Akdemir, and Ergun 2009). It has efficient pollution control with relatively low operating costs and does not require secondary treatment; thus, it is considered a suitable technology for reduce emissions generated during composting (Janni, Jacobson, and Hetchler 2014). Biofilter performance depends on factors, such as media type, porosity, pH, and moisture content (Akdeniz et al. 2011a; Chen, Hoff, and Cai 2009; Scheutz et al. 2011). Another factor, empty bed retention time (EBRT) is a key factor in estimating biofilter performance because it affects microorganism absorption and conversion process. Biofilter performance reported in studies on NH₃ and VOCs removal is shown in Table 1.

Table 1. Comparison of biofilter performance reported in some studies on NH₃ and volatile organic compounds (VOCs) removal

Pollutant	Biofilter media	Empty bed retention time (EBRT, s)	Inlet load (g. m^−3. h^−1)/^a (mg. m^−3. h^−1)	Removal efficiency (%)	Reference
NH3	Organic fraction of municipal solid wastes mixed with chopped pruning waste	86	0.85–7.5	95.9–98.8	Pagans, Font, and Sánchez (2005)
NH3	Digested wastewater sludge mixed with chopped pruning waste	86	6.67	99.4	Pagans, Font, and Sánchez (2005)
NH3	Animal by-products mixed with chopped pruning waste	86	21.7–67.1	46.7–89.5	Pagans, Font, and Sánchez (2005)
NH3	Compost	18–60	11.5	97–99	Yin and Xu (2009)
NH3	Sludge	30–60	9.4	95–99	Yin and Xu (2009)
NH3	Mixture of wood chips, straw and compost	30	0.8–2.6	85–98	Lee et al. (2013)
NH3	Mixture of compost and woodchip	34	5.24	35–80	Yang et al. (2014)
NH3	Mixture of high-density polyethylene and inorganic powders (e.g., activated carbon, zeolite, clay and slag)	60–100	2.7	2–98	Park, Shin, and Chung (2008)
Dimethyl sulfide (DMS)	Mixture of mature compost and yard waste	58.4	34 a	10–100	Hort et al. (2013)
Dimethyl sulfide (DMS)	Mixture of mature compost with sewage sludge and yard waste	58.4	34 a	13–100	Hort et al. (2013)
Dimethyl sulfide (DMS)	Wood bark	14–113	0.833–1.5	65–100	Smet et al. (1996a)
Dimethyl sulfide (DMS)	Compost	31	20.8	100	Smet et al. (1996a)
Dimethyl disulfide (DMDS)	Wood bark	14–113	0.5–1	65–100	Smet et al. (1996a)
Trimethylamine (TMA)	Activated sludge	319	2	90	Chang et al. (2004)
Trimethylamine (TMA)	Granular activatedcarbon	30–60	2	92	Ho et al. (2008)
Trimethylamine (TMA)	Compost	60	9.31	80–100	Ding et al. (2007)
Trimethylamine (TMA)	Sludge	60	9.13	61–100	Ding et al. (2007)
Methanethiol	Mixture of mature compost and yard waste	58.4	7.9 a	95–100	Hort et al. (2013)
Methanethiol	Mixture of mature compost with sewage sludge and yard waste	58.4	7.9 a	70–100	Hort et al. (2013)
Toluene	Wood charcoal	66–264	3.1–16	10–90	Zamir, Halladj, and Nasernejad (2011)
Toluene	Mixture of compost and beads	71–120	15.5–114.3	82–97	Rajamanickam and Baskaran (2017)
Mixture with benzene, toluene, ethylbenzene and xylenes	Mixture of scoria and compost	83–138	/	66–98	Amin et al. (2017)

Notes. a: mg. m⁻³. h⁻¹;/: no data.

Largely different gas removal efficiencies were reported despite using the same EBRT for biofilters. In addition, NH₃ and VOCs emissions from dead animals composting were different from those generated by composting without carcasses because of the decomposition of animal tissues during composting (Akdeniz, Janni, and Salnikov 2011b; Glanville et al. 2016; Thomson and Heyst 2008).The gas species may behave differently in the biofilter, especially when varieties of different gas components that interact with each other were present simultaneously (Ferdowsi et al. 2017). The aims of the present research are (1) to identify the main VOCs contributing to odors during composting of dead pigs and manure based on odor active values (OAVs); (2) to study the effect of EBRT on the removal efficiencies (REs) of NH₃ and VOCs; and (3) to evaluate the elimination capacity (EC) of NH₃ and VOCs by using mature compost as filter media. The results will be useful for the utilization of EBRT in full-scale biofilter operations.

Experiments and methods

Composting of dead pigs and manure

The fresh pig manure and dead pigs used for composting in this study were collected from a pig farm in Beijing, China. Corn stalks were obtained from Anding Town, Daxing District, Beijing, China. The corn stalks (chopped to 1–5 cm length) were used as bulking agents for composting, and the volume ratio between pig manure and the cornstalks was set at 1:2 to achieve a mixture bulk density of about 600 kg m³. The characteristics of pig manure and corn stalk used for composting are shown in Table 2.

Table 2. Initial characteristics of composting materials

Material	pH	Water content (%)	Carbon content (%) (dry basis)	Nitrogen content (%) (dry basis)
Pig manure	6.8	78.2 ± 1.5	33.5 ± 1.6	3.4 ± 0.6
Corn stalk	n.d.	14.6 ± 1.7	39.2 ± 1.6	1.8 ± 0.5
Mixture	7.5	69.4 ± 2.6	34.4 ± 1.2	3.0 ± 0.4

Notes. n.d.: not detected.

Experimental setup

The schematic of deodorization system is shown in Figure 1.

Figure 1. Schematic of the composting and biofiltration system. (1) Fermentation tank; (2) air distribution pipe; (3) blower; (4) odor collection pipe; (5) deodorization fan; (6) valve, (7) mass flow meter; (8) flow meter; (9) biofilters; gas sampling point (10) from composting, (11) after biofiltration, and (12) in the air; (13) temperature loggers; (14) multichannel sampler; (15) vacuum pump; and (16) gas monitor

The dead pigs and pig manure were composted in six thermally insulated 0.95 m³ reactors made of polyvinyl chloride (PVC). The external dimensions of each reactor were 1 m in length, width and height. The mixtures of pig manure and chopped corn stalks were placed on the bottom of each reactor until the mixtures were 40 cm in depth. The dead pigs were then placed on top of each mixture, and 20 cm layers of the mixtures were used to cover the carcasses. Air was supplied to every reactor intermittently with a blower (2PB710-H37, FengLiShi Instrument Technology Co., Ltd., Suzhou, China). The aeration blowers were operated intermittently with on/off times of 10/20 min. The air flow rate was precisely controlled and measured by a thermal mass flow meter (RK100, KeLiBoo Instrument Technology Co., Ltd., Beijing, China) and glass rotameter (ZLB-60, TianHu Instrument Technology Co., Ltd., Shanghai, China), ensuring a permanent air flow rate of 100 L min⁻¹ (m⁻³ pile) for every composting reactor. The composting piles were turned and mixed when the temperature decreased to below 45 °C. In terms of forced aeration, the moisture content of the composting mixture decreased as composting process. Therefore, sufficient water was added to the mixture during mixing and the moisture content of the mixture was maintained at 60%.

The deodorization fan (2PB610-H16, FengLiShi Instrument Technology Co., Ltd., Suzhou, China) extracted the air through the compost mass and discharged the exhaust gas to the nine biofilters filled with mature compost as a biofilter media. Each biofilter run was performed with the same media. The gas flow rate in each biofilter was precisely controlled and measured by thermal mass flow meter (RK100, KeLiBoo Instrument Technology Co., Ltd., Beijing, China) and glass rotameter (ZLB-60, TianHu Instrument Technology Co., Ltd., Shanghai, China). The details of these experiments are described in Table 3.

Table 3. Experimental arrangement of biofilters

Treatment	Biofilter	Experimental arrangement
1	Biofilters 1–3	Influent gas flow rate = 8.478 m^3.h^−1, EBRT = 30 s
2	Biofilters 4–6	Influent gas flow rate = 4.239 m^3.h^−1, EBRT = 60 s
3	Biofilters 7–9	Influent gas flow rate = 2.543 m^3.h^−1, EBRT = 100 s

The biofilters were constructed with circular unplasticized PVC pipes, and the dimensions were 1.2 m (height) and 0.3 m (inner diameter). The media depth was 1.0 m. The mature compost, which was composted for 3–4 months, was used as the media for odor removal. The initial and final properties of the mature compost used as the biofilter media are shown in Table 4.

Table 4. Initial and final characteristics of biofilter media

	Water content (%)	pH	Carbon content (% dry basis)	Nitrogen content (% dry basis)	NH4^+-N (% dry basis)
Initial	57.4 ± 2.5	6.9	30.8 ± 2.8	2.9 ± 0.7	0.26
Final
Biofilters 1–3	60.2 ± 3.3	6.8	25.7 ± 4.4	3.1 ± 0.6	0.55
Biofilters 4–6	61.5 ± 4.9	6.7	23.6 ±3.8	3.3 ± 0.5	0.61
Biofilters 7–9	61.6 ±5.5	6.7	23.1 ±4.0	3.4 ± 0.6	0.63

Sampling and analytical method

NH₃ concentrations were analyzed with a photoacoustic multi-gas analyzer (model Innova 1412i, LumaSense Technologies, Ballerup, Denmark). The detection range was 0.2 to 2000 mg m⁻³. Each air sample was analyzed for six cycles (2 min/cycle) for each measurement, with the first five cycles for stabilization and the last cycle for analysis. The gas sampling ports of the biofilter inlets (Figure 1) were located near the inlets of the deodorization fan. Biofilter outlet gas was sampled in the sampling port (Figure 1) through a cross-type tube for keeping the sampling gases representative. A total of 13 sampling ports (three inlet gas, nine outlet gas, and one outside air sampling ports) were connected to a multichannel gas sampling manifold, which could automatically switch from one sampling port to another. Gas measurement for each sampling port was conducted online daily. The analyzers were calibrated with standard calibration gases prior to measurement and regularly checked with zero and span gases for ensuring that the deviations in the measured readings were within 5% of the reference values.

In addition, VOCs sampling and measurement were performed at the sampling port of the outlet of composting fermentation on days 0, 1, 2, 3, 4, 5, 6, 7, 8, 10, 12, 14, 17, 21, 24, 28, 31, 35, and 42 after the start of the test, and the gas discharge outlet of the biofilters was measured on days 0, 7, 10, 14, 21, 28, 35, and 42.

The SUMMA canisters were used to collect air for VOCs determination through the U.S. Environmental Protection Agency (EPA) TO-14 method. Gas from the SUMMA canisters was injected into the cold trap concentrator (ENTECH 7100, USA) for O₂, N₂, and CO₂ removal. After being enriched, the gas was injected to a gas chromatography mass spectrometer (7890A/5975 C) equipped with a chromatographic column (Agilent Technologies, DB-624, 60 m × 0.25 mm × 1.8 μm). The column temperature was initially set at – 10 °C, then maintained for 10 min and then increased to 100 °C at 3.0  °C min⁻¹ and to 220 °C at 10 °C min⁻¹, and finally maintained for 15 min. The injector temperature was 140 °C, the solvent delay time was 0.5 min, and the carrier gas flow rate was 1.0 mL min⁻¹. The ion source temperature was 250 °C, and the scanning mode was selection scanning (SIM). The quantification of the total VOCs concentration was based on the retention time and peak area. Two kinds of standard gases (US EPA PAMS and US EPA TO-14), including 35 alkanes, 22 aromatic hydrocarbons, 20 halo hydrocarbons, one amine, three sulfur compounds, and four freon were used to quantify the VOCs concentration. The total VOCs (TVOCs) detection limit of the method was 2.8 × 10⁻⁴ to 7.5 × 10⁻³mg m⁻³, whereas the gas injection volume was 400 mL. The parallel analysis concentration was 5 nmol/mol mixed with standard sample 10 times, and the relative deviation was less than 15%.

The temperatures inside the composting piles and outdoor were monitored continuously using HOBO Pro Series Temp data loggers (Onset Computer Corp., Bourne, MA, USA). All temperatures were recorded with interval of 1 h.

The biofilter packing materials were sampled before and after the trials. The pH, nitrogen content (TN), carbon content (TC), water content (WC), and NH₄⁺-N of all the samples and composting raw materials were determined. The pH was measured by a calibrated portal pH meter (PHB-4, Shanghai INESA Scientific Instrument Co., Ltd., China) in 1:10 (m/V) material water extract and defined as the pH concentration. NH₄⁺-N concentrations in the solutions extracted from the solid samples by using 2 M KCl (using a fresh solid sample to KCl weight-to-volume ratio of 1:10) were measured through the TAN-salicylate method 10,031 (0–50.0 mg L⁻¹) as described by Wang, Dong, and Zhu (2016). A Hach DR5000 (Hach Company, Loveland, CO) spectrophotometer was used in the measurements. The WC was determined by weight loss through oven drying at 105 °C for 24 hr. TC and TN contents were determined by analyzing each dried and ground sample in triplicate with an elemental analyzer (PE 2400, PerkinElmer, USA).

Statistical analyses

The gas average RE (%) is calculated using Equation (1)(1) $RE=\frac{C_{IN}-C_{OU}}{C_{IN}}\times 100\mathrm{\%}$(1) .

(1) $RE=\frac{C_{IN}-C_{OU}}{C_{IN}}\times 100\mathrm{\%}$(1)

where C_IN and C_OU are the inlet and outlet gas concentrations (mg m⁻³), respectively.

The temperature, inlet gas concentrations, and gas REs were analyzed by repeated measure analysis of variance on Excel 2016.

Results and discussion

Temperature of composting of dead pigs and manure

Temperature is an important index of composting. During composting, the outside daily temperature was between – 5.6 °C and 14.3 °C. Microorganisms absorbed small degradable organic molecules by catabolism and released large amounts of heat in the initial stage (Zhu et al. 2016), which induced the temperature of six composting piles to increase significantly to 60.3 °C, 53.1 °C, 60.4 °C, 57.6 °C, 38.4 °C, and 60.4 °C on day 2, respectively. The piles were turned when the temperature was lower than 45 °C, and all the six composting piles were turned on day 30 of composting (Figure 2). Maintaining an average temperature of 55–60 °C for a couple of days is sufficient to eliminate nearly all pathogenic viruses, bacteria, fungi, and protozoa (Kalbasi et al. 2005; Xu et al. 2010). Compost temperatures of 40–50 °C are sufficient to eliminate avian influenza and Newcastle disease viruses in less than 3 days (Guan et al. 2009). Highly pathogenic avian influenza viruses present in fresh animal manure can be inactivated at 35 °C within the 24 h of composting (Elving et al. 2012). In the present study, the daily average temperature, maintained > 55 °C for at least 12 days, and thus met the hygienic requirements for harmless disposal in GB7959-2012 (Chinese Standard 2012) and technical requirements for nonhazardous treatment of animal manures in NY/T1168-2006 (Chinese Standard 2006).

Figure 2. Temperature of composting piles (arrow means turning)

The main odors VOCs components

Odors are attributed to the interactions, such as antagonism and synergism, of various stinky substances (Blazy et al. 2014). Olfactometry analysis is recommended and extensively used in assessing odor (Zhu et al. 2016). The dynamic olfactometry and triangle odor bag methods are the most common olfactometry approaches (Lu et al. 2015). However, they are quite expensive and do not identify odor concentrations for individual odorous pollutants within a sample (Han et al. 2019).

OAV is widely used in odor assessment due to its low cost for identifying odor contributions for individual chemicals within a sample (Anfruns, Martin, and Montes-Morán 2011; Sivret et al. 2016; Zhang et al. 2016b, 2016c). It is calculated using Equation (2)(2) $OAV=\frac{C_i}{C_{io}}$(2) .

(2) $OAV=\frac{C_i}{C_{io}}$(2)

where C_i is the odor in ambient air (mg·m⁻³), and C_io is the olfactory threshold of the odor (mg·m⁻³).

The gas chromatography mass spectrometry analysis of the different samples during composting yielded 38 pollutants, including three volatile organic sulfur compounds (VOSCs), 11 aromatic hydrocarbons, 23 haloalkanes, and one ketone. Table 5 showed the compounds and their concentrations, odor thresholds (OTs), and OAV of the individual chemicals of VOCs.

Table 5. Gas compound concentrations of volatile organic compounds during dead pigs and manure composting

Component	Detecti-on limits (mg.m^−3)	Compost time (d)																				OT (mg.m^−3)	OAVs	OT References
		0	1	2	3	4	5	6	7	8	10	12	14	17	21	24	28	31	35	42	AV
Benzene	0.0034	9.32	48.6	51.0	28.8	17.9	48.5	33.4	31.1	42.6	30.2	22.9	22.9	10.9	4.67	8.88	3.44	18.1	0.00	80.9	27.0	2.7	0.01	Nagata and Takeuchi (1990)
Toluene	0.0026	554	32.5	259	219	92.1	252	79.0	48.7	2063	45.2	81.8	81.8	14.8	5.23	6.25	5.12	3.49	108	15.0	209	0.38	0.55	Cometto-Muniz and Abraham (2015)
Ethylbenzene	0.0005	26.1	56.0	149	125	25.5	126	54.7	25.9	24.9	27.1	40.3	40.3	1.14	-	1.30	-	20.4	1.51	7.19	44.2	0.17	0.26	Van Gemert (2011)
Chlorobenzene		5.86	2.68	3.08	4.57		1.58	3.00	1.26	-	1.43	1.54	1.54	0.00	-	-	-	-	4.80	-	2.6	0.98	0.003	Ruth (1986)
Naphthalene	0.0014	377	64.1	846	1711	1.39	1048	-	166	-	-	-	-	-	-	134	-	-	-	-	544	0.058	9.4	Ruth (1986)
Styrene	0.0076	109	216	254	277	73.0	272	189	71.9	9.86	108	111	111	10.3	-	9.63	9.85	7.57	11.0	10.7	103.	0.15	0.689	Nagata and Takeuchi (1990)
Orthodoxxylene	0.030	12.5	35.4	102	62.5	21.5	79.4	14.3	11.2	16.1	7.50	39.5	39.5	-	-	-	-	15.7	24.6	25.2	33.8	N-D	N-D
Meta xylene and P-xylene	0.0017	12.6	42.7	144	77.2	25.8	95.3	14.8	13.3	20.8	9.87	47.4	47.4	1.71	-	0.00	0.00	19.0	2.98	44.1	34.4	12	0.003	Ruth (1986)
1,3-Dichlorobenzene	0.0003	-	3.45	-	-	4.12	1.88	-	4.37	0.27	-	18.4	18.4	-	-	-	-	-	-	-	7.3	12	0.0006	Ruth (1986)
1,4-Dichlorobenzene	0.0009	-	0.92	-	-	1.62	-	-	1.88	-	-	-	-	-	-	-	-	-	-	-	1.5	12	0.0001	Ruth (1986)
1,2,4-Trichlorobenzene	0.0009	10.2	-	84.7	78.3	-	13.9	-	-	-	0.91	13.7	13.7	-	-	-	-	18.7	17.7	1.70	25.4	N-D	N-D
Difluoro dichloromethane	0.0063	3.94	8.11	8.52	16.0	-	-	6.63	26.9	9.84	10.3	9.08	9.08	3.49	4.43	4.60	4.11	-	-	7.30	8.8	N-D	N-D
1,2-dibromoethane	0.0014	-	-	-	-	-	-	0.00	10.3	-	-	-	-	-	-	-	-	-	-	10.3	6.9	N-D	N-D
Vinyl chloride	0.0011	14.1	16.4	10.4	22.9	-	13.2	21.1	76.2	2.85	16.2	8.74	8.74	4.66	1.87	5.46	2.73	4.26	1.72	9.19	13.4	N-D	N-D
Methyl bromide	0.0013	1.30	0.00	1.47	1.78	-	1.54	-	0.00	2.37	1.86	-	-	4.28	1.50	1.44	2.29	-	-	-	1.7	80	0.00002	Ruth (1986)
Chloroethane	0.0084	3986	8.44	-	-	13.5	-	-	86.6	32.9	26.5	-	-	16.1	23.7	-	43.5	-	-	-	471	N-D	N-D
Trichlorofluoromethane	0.0012	-	-	7.50	9.79	2.05	15.3	-	10.3	3.83	5.25	-	-	-	-	-	-	-	-	-	7.7	N-D	N-D
1,1-dichloroethylene	0.0004	-	-	-	-	-	-	-	-	-	0.43	-	-	-	-	0.44	-	-	-	-	0.4	N-D	N-D
Trichlorotrifluoroethan	0.0012	-	-	-	3.61	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	3.6	0.34	0.001	Ruth (1986)
Dichloromethane	0.0005	9.66	7.14	22.4	156	47.5	61.3	22.8	526	174	105	13.7	13.7	35.4	38.5	24.0	40.4	248	7.63	214	93.0	160	0.001	Nagata and Takeuchi (1990)
1,1-dichloroethane	0.0004	-	-	0.80	0.57	0.48	0.54	-	-	37.2	11.7	-	-	-	-	-	-	-	-	-	8.6	N-D	N-D
Cis-1,2-dichloroethylene	0.0014	1.71	1.37	1.72	1.43	1.37	1.42	-	-	-	-	-	-	-	-	-	-	-	-	-	1.5	0.34	0.004	Ruth (1986)
Trichloromethane	0.0012	6.67	35.5	65.9	19.6	20.6	19.3	3.17	2.15	1.38	6.04	4.38	4.38	1.67	0.00	3.92	3.41	63.8	11.7	27.6	15.9	N-D	N-D
Carbon tetrachloride	0.0011	-	3.79	5.19	2.19	4.50	4.21	2.85	1.28	-	2.06	1.91	1.91	-	-	-	-	-	-	-	3.0	4.6	0.001	Nagata and Takeuchi (1990)
1,2-dichloroethane	0.0009	33.7	259	395	67.8	32.9	44.5	13.8	14.4	-	10.6	7.97	7.97	1.02	1.29	3.71	1.66	27.9	11.6	45.1	54.4	N-D	N-D
Trichloroethylene	0.0003	1.46	4.23	3.35	0.70	1.61	5.48	3.72	2.30	1.07	2.05	0.73	0.73	-	-	0.33		3.04	1.17	0.58	2.0	3.9	0.001	Nagata and Takeuchi (1990)
1,2-dichloropropane	0.0006	6.06	35.5	61.7	30.2	11.1	52.3	3.37	9.80	4.55	4.17	9.85	9.85	0.61	0.67	0.90	0.83	23.4	20.8	12.9	15.7	N-D	N-D
Trans 1,3-dichloropropene	0.0011	-	1.51	-	1.78	1.12	2.09	1.82	1.07	-	2.16	3.87	3.87	-	-	–	-	-	-	-	2.1	N-D	N-D
CIS 1,3-dichloropropene	0.0013	78.8	5.09	20.7	-	1.39	9.23	8.38	3.27	1.78	4.12	9.01	9.01	-	-	-	-	3.63	7.46	3.09	11.8	N-D	N-D
1,1,1-trichloroethane	0.0068	-	-	-	-	-	-	-	-	6.77	-	-	-	-	-	-	-	-	-	-	6.8	N-D	N-D
1,1,2-trichloroethane	0.0012	86.5	3.05	66.4	11.6	9.03	24.7	20.6	3.75	24.4	4.95	41.3	41.3	0.00	0.00	16.9	0.00	3.94	72.8	1.23	22.8	N-D	N-D
Tetrachloroethylene	0.0021	48.0	6.34	38.7	32.0	9.49	68.2	-	15.4	-	11.7	-	-	-	-	-	-	3.04	7.11	4.67	22.3	N-D	N-D
1,2-dibromoethane	0.0014	2.01	10.5	6.92	7.27	4.00	2.21	1.87	2.76	1.49	1.80	7.04	7.04	3.46	-	4.01	2.30	-	-	-	4.3	N-D	N-D
Tetrachloroethane	0.0012	35.7	99.6	78.1	184	32.9	73.7	9.25	25.5	911	13.6	16.4	16.4	1.23	-	1.25	2.22	16.5	78.1	10.9	84.6	N-D	N-D
Trimethylamine	0.0063	5729	146	6347	2598	274	-	-	47.0	37.6	0.00	132	132	20.0	27.4	986	0.00	-	698	149	912	0.00003	28,489	Nagata (2003)
Dimethyl sulfide	0.0004	0.00	386	3312	2054	396	1806	670	1883	165	156	0.00	0.00	1748	684	331	155	-	58.1	963	777	0.0076	102.3	Nagata and Takeuchi (1990)
Dimethyl disulfide	0.0004	1100	1784	5065	6540	832	4634	3071	0.00	4.66	376	0.00	0.00	878	845	1321	98.2	-	194	987	1980	0.0085	233.0	Nagata and Takeuchi (1990)
Dimethyl trisulfide	0.0003	3743	154	2173	2753	-	1781	19.9	1543	3.28	0.00	0.00	0.00	59.6	-	1133	58.6	-	417	644	762	0.0087	87.6	Nagata and Takeuchi (1990)

Notes. –Data below detection limit; AV: average of individual chemical concentration during the whole composting period;OT: Olfactory threshold; N-D: not data found; The unit of VOCs from the composting is μg.m⁻³.

The average OAV was calculated using the average of individual chemical concentration during the whole composting period divided by OTs. The OAV of TMA was 28,489 (Table 5), which implied an extremely high odor concentration in the air. The average OAV of individual VOSCs were ranked as follows: DMDS (233) > DMS (102) > DMTS (87), see Table 5 for information. Thus, TMA, DMDS, DMS, and DMTS were the main targets of the treatment for dead pigs and manure composting. The VOSCs were responsible for 80%–90% of odors emitted during the composting of sewage sludge. This nuisance was mainly due to the extremely low OT values of these products (Hort et al. 2009). In this study, these compounds were monitored due to the important quantities of the emitted NH₃ and VOSCs.

NH₃ and VOCs RE of biofilter

NH₃ removal efficiency

The daily average NH₃ concentrations in the inlet gases of the biofilters was 12.7–446.8 mg.m⁻³, and the NH₃ concentrations in the outlet gas of treatments 1 (EBRT = 30 s), 2 (EBRT = 60 s), and 3 (EBRT = 100 s) were 2.4–60.8, 3.3–54.1, and 4.2–56.9 mg m⁻³, respectively (Figure 3). On the first day of composting, the daily average NH₃ concentration was 446.8 mg m⁻³, which was the highest value. By contrast, the NH₃ concentrations in the outlet gas decreased to 53.1, 38.7, and 30.7 mg m⁻³ in treatments 1, 2, and 3, respectively, with corresponding REs of 88.1%, 91.3%, and 93.1% compared with the concentrations in the inlet gas. The average REs of treatments 2 and 3 were kept at nearly constant values, that were high than 85% from days 2 to 10, and were 74.6% and 74.5%, respectively. At day 12, the RE reached 90% again and were 83%–93% and 82%–92% until day 30. It reached 98% after the mixing of all the piles. The RE eventually declined but remained above 75% until the end of experiments. After 30 days of composting, the average ER of treatment 1 was 82.4%, which was lower than the values in treatments 2 and 3. However, from day 30 to day 42, the average RE of treatment 1 increased to 91.2%, which was higher than the values of treatments 2 and 3. In the entire composting period, the average REs for treatments 1, 2, and 3 were 85.4%, 88.7%, and 89.0%, respectively. No significant difference was found between the treatment efficiencies of treatments 2 and 3, but these rates were significantly higher than the rate of treatment 1. Many studies have utilized mature compost with/without other materials as packing media to remove NH₃ emissions from the composting materials and obtained a total NH₃ RE of 35%–99% (Hong and Park 2004Lee et al. 2013; Pagans, Font, and Sánchez 2005; Yang et al. 2014; Yin and Xu 2009). Some differences in REs are present among the studies because RE is affected by many factors, such as compost source, pH, oxygen concentration, and nutrients in biofiltration systems (Hong and Park 2004; Pagans, Font, and Sánchez 2005; Wu et al. 2010). In some studies, inoculating external microorganisms in biofilters can shorten or even eliminate the startup phase, thereby resulting in high gas REs within the first 2 days (Park et al. 2013). In the present study, no start-up period was observed for NH₃ removal due to the high NH₃ adsorption and absorption capacities of the mature compost media and sufficient potential nitrification activity for the oxidization of the average amount of NH₃-N that entered the biofilter per day.

Figure 3. NH₃ concentration inlet and outlet gas of biofilter and average RE of treatments with different content times

In this study, the RE was higher than 70% in all the treatments, except in days 10, 11, and 42. Two composting piles were turned and mixed on day 10, and the experiment ended on day 42. However, the ammonia analyzer was still sampling and measuring during the turning of composting, which may result to low REs.

VOCs removal efficiency

The components of VOCs emitted from the composting of dead pigs were complicated. The effects of EBRT on the RE of DMS, DMDS, DMTS, and TMA were analyzed (Figure 4). The VOCs were effectively removed by the biofilters, but the differences in RE among the VOCs were still observed. TMA presented the highest RE, followed by DMTD, DMDS, and DMS.

Figure 4. TVOCs inlet and outlet gas of biofilter and average RE for treatments with different content times of (a) DMS, (b) DMDS, (c) DMTS, and (d) TMA

At the beginning of composting, the TMA concentration was relatively high at 6 mg.m⁻³. The concentration after biofilters decreased to 0–0.7 mg m⁻³. Turning the composting materials caused the rise of concentration to 4 mg m⁻³. The average REs of TMA for treatments 1, 2, and 3 were 92.9%, 100%, and 100%, respectively. The highest concentrations of DMS, DMDS, and DMTS were detected on the first day of composting (2.65, 11.0, and 2.74 mg m⁻³, respectively). Afterward, the concentration began to decrease slowly. The fluctuations in concentration were caused by turning. The average REs of DMS were 79.2% (treatment 1), 82.2% (treatment 2), and 95.4% (treatment 3). The average REs of DMDS were 81.9% (treatment 1), 89.1% (treatment 2), and 94.0% (treatment 3). The REs of DMTS were 76.7% (treatment 1), 98.4% (treatment 2), and 99.1% (treatment 3).

Hort et al. (2013) used a mixture of mature compost and yard waste or sewage sludge as packing media to treat DMS and the RE was 10%–100%. Smet et al. (1996a, 1996b) obtained DMS and DMDS REs of 65%–100%. Ho et al. (2008) demonstrated that TMA was more difficult to degrade than other amines, such as MA and DMA. In the present study, an RE of approximately 92%–100% for TMA was achieved. The optimal pH in the biofilter for TMA removal ranged from 6.0 to 8.0, and excellent results were obtained at pH 7.0 (Ho et al. 2008). The pH of the mature compost used for the biofilter media was 6.7–6.8, which can result in maximal enzyme activity for TMA degradation by Paracoccus sp. CP2 (Chang et al. 2004), the microorganism that is often used to degrade TMA.

The REs of the four VOCs in treatments 2 and 3 were significantly higher than that in treatment 1. The decrease in EBRT usually leads to lower RE due to the insufficient contact time between the pollutant and microbial population (Fu et al. 2011; Kim, Cai, and Sorial 2005). The present study considered the EBRT of 60 s as suitable for VOCs removal from composting of dead pigs and pig manure.

Elimination capacities of NH₃ and VOCs

Elimination capacity of NH₃

Biofiltration performance was evaluated in terms of elimination capacity (EC) to reflect the capacity to remove pollutants. The EC of NH₃ was plotted as a function of the inlet load (IL) in Figure 5a. The EC of IL increased gradually at a slow rate. In this study, the maximum IL was only 53.2 g m⁻³ h⁻¹, corresponding to the maximum EC of 47.2 g m⁻³ h⁻¹. The EC maximum of the biofilters packed with mature compost was not observed. High NH₃ REs are usually obtained at high NH₃ concentrations because of the adsorption and absorption capacities of the media, but performance declines when NH₃ load exceeds those capacities and biological activity (Pagans, Font, and Sánchez 2005). The maximum safe NH₃ loading depends on nitrification activity (Yasuda et al. 2009). In the present study, the IL was lower than that in another study by Ryu, Cho, and Lee (2011), in which the EC of NH₃ at 149 g m⁻³ h⁻¹ was maintained during the 125-day operation period. Mohammad, Veiga, and Kennes (2007) and Xi, Hu, and Qian (2006) found that if the inlet concentrations are lower than those under saturation condition, the increasing inlet concentrations can enhance the transfer rates of pollutants. These results were also obtained in the present study (Figure 5b). The EC increased with inlet concentration, and the increased rate in treatment 1 (with 30 s) was higher than in treatments 2 and 3 because of the higher IL in treatment 1.

Figure 5. (a) Variations in removal efficiency and elimination capacity of NH₃ with changes in inlet load and (b) influence of inlet concentration on RE and EC (The circle, EBRT 30s; triangle EBRT60s; cross, EBRT 90s)

Elimination of VOCs

The maximum ECs of DMS, DMDS, DMTS, and TMA were 220, 1301, 296, and 603 mg m⁻³ h⁻¹, respectively (Figure 6a). Hort et al. (2013) used yard waste as filter media for sulfur compound (RSC) treatment and showed that the maximum EC of these RSCs was 935 mg m⁻³ h⁻¹. Smet et al. (1996a, 1996b) found that the maximum EC of DMS and DMDS were 0.8–20.8 and 0.5–1 g m⁻³ h⁻¹ when using wood bark and compost as the media. Figure 6a indicates that the EC of DMS, DMDS, DMTS, and TMA increased with the increased of IL. EC of DMDS was most of the time similar to the inlet load. This implied that the RE of DMDS was relatively high, close to 100% (Figure 6b). On the contrary, the EC of DMS, DMTS and TMA, were often different from the IL. This is inversely proportional to the solubility of the VOCs studied, where DMDS is the less soluble compound than other three kinds of VOCs (Table 6). The RE of biofilter is negatively correlated to the IL of VOCs. The inhibition of microbial activity by some VOCs in high IL may be responsible for the decrease in RE (Fu et al. 2011). The RE decreased with the increase of DMS, DMTS, and TMA throughout the whole experimental period. The RE of DMDS was different from that of the other VOCs due to insolubility of DMDS in water. Also, some components in the composting exhaust gases had an effect on VOCs removal (Hort et al. 2013), but further research is required to investigate these results.

Table 6. The solubility of DMS, DMDS, DMTS, and TMA

VOCs	Solubility in water
DMS	Slightly soluble in water;	2.2 * 10^4 mg/L at 25 °C
DMDS	Insoluble in water	3*10 ^3 mg/L at 25 °C
DMTS	Very slightly soluble in water	n.d.
TMA	Soluble in water	8.9 * 10^5 mg/L at 30 °C

n.d. means no date.

Figure 6. (a) Elimination capacity and (b) removal efficiency of DMS, DMDS, DMTS, and TMA versus inlet load

Physicochemical properties of biofilter media

Mature compost can provide large surface area and permeable pore space, which are suitable for microbe proliferation (Scheutz et al. 2011). It can also contain several natural microbial communities that can biodegrade various pollutants; thus, compost media-based biofilters have been extensively used for NH₃ and VOCs removal (Datta and Philip 2014; Pagans, Font, and Antoni 2006; Schnelle, Dunn, and Ternes 2015). WC is a parameter that plays an essential role in biofiltration because microorganisms require water to maintain their normal metabolic activities. The optimal range of relative humidity for organic and biologically active media is between 30% and 60% (Mudliar et al. 2010). In the present study, the initial WC of the media was 57.4%, and low variations were found during experiment. The pH of the biofilter media did not fluctuate and ranged from 6.7 to 6.9. Yasuda et al. (2009) found that nitrification in biofilter can increase the pH of a biofilter. As a result, the packing materials had neutral pH levels in this biofiltration system. Moreover, the pH levels remained nearly constant (close to 7) during the entire duration of the experiment.

Conclusion

A total of 37 kinds of VOCs present in the air were identified, and temporal variations were determined during 42 days of dead pigs and manure composting. DMS, DMDS, DMTS, and TMA were identified as the main pollutions of VOCs components based on their OAVs. The inlet concentration of NH₃ applied was 12–447 mg m⁻³, and the average REs were 85.4%, 88.7%, and 89.0% for EBRTs of 30, 60, and 100 s, respectively. The average REs of DMS, DMDS, DMTS, and TMA were 79.2%–95.4%, 81.9%–94.0%, 76.7%–99.1%, and 92.9%–100%, respectively, and their maximum ECs were 220, 1301, 296, and 603 mg m⁻³ h⁻¹, respectively. Overall, this study suggested an EBRT of 60 s for the stimulation removal of NH₃, DMS, DMDS, DMTS, and TMA.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work was supported by the National Key Research and Development Program of China [2016YFD0501406] and China Agriculture System [CARS-36].

Notes on contributors

Bin Shang

Bin Shang is a Associate Professor, Institute of Environmental and Sustainable Development in Agriculture (IEDA). Chinese Academy of Agriculture Sciences (CAAS), Address: 12 Zhongguancun South Street, Beijing 100081, P.R. China, Email: [email protected], Tel/Fax: [phone]. Research extension is about Livestock Environment Control and Waste Treatment.

Tanlong Zhou

Tanlong Zhou is a Doctoral candidate, Institute of Environmental and Sustainable Development in Agriculture (IEDA). Chinese Academy of Agriculture Sciences (CAAS), Address: 12 Zhongguancun South Street, Beijing 100081, P.R. China, Email: [email protected], Tel/Fax: [phone].

Xiuping Tao

Xiuping Tao is a Professor, Institute of Environmental and Sustainable Development in Agriculture (IEDA), Chinese Academy of Agriculture Sciences (CAAS), Address: 12 Zhongguancun South Street, Beijing 100081, P.R. China. Email: [email protected], Tel/Fax: [phone]. Research extension is about livestock environment control and waste treatment.

Yongxing Chen

Yongxing Chen is a Associate Professor, Institute of Environmental and Sustainable Development in Agriculture (IEDA). Chinese Academy of Agriculture Sciences (CAAS), Address: 12 Zhongguancun South Street, Beijing 100081, P.R. China, Email: [email protected]. Tel/Fax: [phone]. Research extension is about livestock waste testing.

Hongmin Dong

Hongmin Dong is a Professor, Deputy Director, Institute of Environmental and Sustainable Development in Agriculture (IEDA), Chinese Academy of Agriculture Sciences (CAAS), Address: 12 Zhongguancun South Street, Beijing 100081, P.R. China, Email: [email protected], Tel/Fax: [phone]. Research extension is about livestock environment control and waste treatment.