Quantification of CH₄ and N₂O Fluxes from Piggery Wastewater Treatment System for Emission Factor Development

Abstract

Piggery farming is the largest source of livestock manure in South Korea, yet greenhouse gas (GHG) data from piggery wastewater treatment systems remain limited. This study quantified methane (CH₄) and nitrous oxide (N₂O) fluxes from a full-scale treatment facility to develop stage-, seasonal-, and diurnal-specific emission factors. Continuous laser-based monitoring using a PVC air-pool chamber was applied across raw wastewater storage, an anoxic nitrogen-conversion reactor, and strongly aerated nitrification units. Mean CH₄ fluxes ranged from 1.1 to 15.6 mg s⁻¹ m⁻² peaking in summer, while N₂O fluxes ranged from 0.01 to 17,971 mg s⁻¹ m⁻², with maxima in fall. Emissions were dominated by two functional zones: aerated basins where vigorous mixing enhanced CH₄ stripping, and an upstream anoxic reactor where oxygen instability and nitrite accumulation produced extreme N₂O peaks. Derived emission factors were 0.11 kg CH₄ head⁻¹ yr⁻¹ and 45.2 kg N₂O head⁻¹ yr⁻¹, equivalent to 3.1 and 12,300 kg CO₂-eq head⁻¹ yr⁻¹. CH₄ variability was controlled mainly by treatment stage and temperature, whereas N₂O was governed by internal redox conditions. These results refine emission factors for inventories and underscore the need for improved aeration stability and denitrification control to reduce GHG emissions from piggery wastewater systems.

1. Introduction

Greenhouse gas (GHG) emissions, particularly methane (CH₄) and nitrous oxide (N₂O), are major contributors to global climate change. Agriculture accounts for approximately 9 to 18% of global anthropogenic GHG emissions, with livestock manure management recognized as a significant source of both CH₄ and N₂O [1]. International frameworks such as the Paris Agreement emphasize the need for accurate GHG quantification to guide mitigation strategies [2], yet emissions from livestock wastewater remain difficult to characterize due to methodological inconsistencies, high spatial and temporal variability, and limited emission factor (EF) datasets [3,4].

South Korea has pledged carbon neutrality by 2050 [5], with livestock-sector mitigation identified as a key priority. As of 2023, the national livestock population reached roughly 190 million head, including 11.09 million pigs [6]. Piggery operations generate the largest share of manure in Korea, contributing about 39% (≈19.7 million tons yr⁻¹) of the national total [7]. Despite this, comprehensive CH₄ and N₂O emission factors for piggery wastewater treatment processes remain scarce, as most previous studies have focused on barn-level or slurry-pit emissions. This gap limits the refinement of national inventories and assessment of mitigation opportunities [8,9].

Quantifying GHG emissions from pig manure systems remains challenging because emissions vary widely with chamber configuration, sampling frequency, environmental conditions, and treatment-stage operation [8,10,11,12]. Flux-gradient and micrometeorological approaches can be difficult to apply in complex wastewater environments such as aeration basins and storage tanks [13,14]. Emissions also respond strongly to operational and seasonal conditions: CH₄ generally increases during warm periods, while N₂O often peaks in cooler, oxygen-limited conditions [13,14]. Existing empirical models only partially capture these dynamics, contributing to uncertainty in current inventories [15,16].

Advances in laser-based gas analyzers and automated monitoring systems now allow high-resolution CH₄ and N₂O measurements with improved stability and sensitivity [17,18,19,20]. However, their application in piggery wastewater treatment systems remains limited, and localized emission factors reflecting diurnal and seasonal variability are lacking [21].

This study quantifies CH₄ and N₂O emissions from a commercial piggery wastewater treatment facility and evaluates how fluxes vary across treatment stages, diurnal periods, and seasons. By deriving stage- and time-resolved emission factors, this work fills a critical data gap in Korea’s GHG inventory framework and provides empirical guidance for designing low-emission wastewater treatment strategies under warm, humid agricultural conditions.

2. Materials and Methods

2.1. Waste Water Treatment Plant and Weather Conditions

The study was conducted at a commercial pig farm located in Baeksan-myeon, Jojong-ri, Kimje City, South Korea (35°52′1.31″ N, 126°52′41.31″ E). Figure 1 illustrates the four-season classification system applied in this study for seasonal flux analysis. Table 1 presents the average monthly weather characteristics of the study area during the monitoring period. The experimental site geographical location and layout of the swine wastewater treatment system used for GHG monitoring are represented in Figure 2. The pig farm houses approximately 9000 pigs, generating an average 16.47 $\mathrm{t} \mathrm{d}\mathrm{a}{\mathrm{y}}^{-1}$ of solid waste and 25.65 $\mathrm{t} \mathrm{d}\mathrm{a}{\mathrm{y}}^{-1}$ of effluent. The treatment plant is a self-purification wastewater treatment system designed based on aerobic and anaerobic processes to degrade organic content at multiple treatment stages (Figure 2B).

Figure 1. Seasonal variations in ambient environmental conditions (temperature, relative humidity, and wind speed) obtained from the nearest Korea Meteorological Administration (KMA) station during the measurement period. Wind speed corresponds to the standard 10 m measurement height.

Table 1. Average annual weather conditions of the study area during the experiment period.

Index	Fall	Winter	Spring	Summer
Wind speed (m/s)	1.2	1.5	2.0	1.7
Temperature (°C)	18.0	1.5	13.4	26.5
Relative humidity (%)	75.3	69.9	64.5	74.9

Figure 2. Experimental site location and layout of the wastewater treatment system used for GHG monitoring. (A) Geographical location of the study site in Korea. (B) Aerial view of the farm and waste treatment facility showing sampling points (1 to10) and key units (a to d). (C) Illustration of the treatment sequence, tank dimensions, and sampling stages.

From Figure 2, the treatment process begins at Tank (a), where raw waste is received before undergoing a separation process at (b) to divide solid and liquid waste fractions. The liquid waste is then stored in liquid waste tanks (c) and (d) before being pumped into the treatment unit. The wastewater first accumulates in an underground storage tank (Stage 1) and then moves to the anaerobic treatment stage (Stage 2), where microorganisms break down organic matter under oxygen-deficient conditions, releasing GHG as a byproduct. Following this, the waste undergoes aerobic treatment (Stages 3 to 7), where organic matter is further degraded in aeration tanks (Aerobic 1–5) under oxygen-rich conditions. Secondary anaerobic treatment occurs in Anaerobic Tank 2 (Stage 8), followed by secondary aerobic treatment in Aerobic Tank 2 (Stage 9) before final discharge.

Figure 2C provides a detailed layout of the wastewater handling process, technical specifications of the treatment unit, and GHG monitoring points considered in this study. The CH₄ and N₂O emissions were monitored at three main stages: pre-treatment (raw waste storage), primary treatment (anaerobic and aerobic stages), and secondary treatment (anaerobic and aerobic stages).

2.2. Equipment and GHGs Collection

2.2.1. Equipment

A custom floating PVC air-pool chamber was installed at Stages 2, 3, 8, and 9 to enable continuous gas capture while preventing corrosion from wastewater emissions (Figure 3). The chamber measured 1.22 m × 2.44 m × 0.60 m (collection area 2.98 m²) and was connected to flexible air ducts, an air blower, and 0.25-inch Teflon tubing for transport to the analyzer (Figure 4a). At Stage 1, where chamber deployment was not possible due to the underground location and restricted access, a stainless-steel cover with two inlet ports and one outlet port was used, and gas was extracted using 117 m³ s⁻¹ blowers (Figure 4b).

Figure 3. Schematic illustration of gas collection and monitoring setup, showing the isometric and cross-sectional views of the PVC air pool system. The setup includes an air blower airflow, Teflon sampling tubes for inlet and outlet gas collection, and a laser gas spectrometer (LGS) unit for real-time CH₄ and N₂O measurement.

Figure 4. Field installation and analytical system used for GHG flux monitoring. (a) Floating PVC air pool chamber deployed at monitoring stages 2, 3, 8, and 9; (b) gas collection and transfer setup at the raw-water storage tank (stage 1) and (c) Laser Gas Analyzer (LGR) unit.

Gas concentrations were measured using a laser-based analyzer (ABB Measurement & Analytics, Quebec City, QC, Canada; LGR-ICOS™ GLA351-N2OM1; Figure 4c). The analyzer alternated between CH₄ and N₂O using its internal optical-switching mechanism. All sampled gas was conveyed through Teflon tubing to minimize adsorption or chemical loss. Sensor detection ranges and measurement methods are summarized in Table 2.

Table 2. Measurement range, interval, and method of the gas sensors.

Sensor Type	Measurement Range	Measurement Cycle	Measurement Method
Methane (CH4)	0–1000 ppm	1 s	Laser analysis
Nitrous oxide (N2O)	0–40 ppm	1 s	Laser analysis

Before each monthly monitoring event, the analyzer was calibrated using certified low- and high-concentration CH₄ and N₂O standards. Calibration checks included zero stability, span accuracy, and spectral-fit performance according to manufacturer guidelines.

2.2.2. Gas Collection

Monitoring was performed for ten 24-h campaigns from 6 September 2024 to 30 August 2025, covering autumn, winter, spring, and summer. During each campaign, continuous sampling was conducted at all five treatment stages. The PVC chamber inlets and outlets were monitored simultaneously, with the inlet measurements serving as the blank reference. The detailed monitoring days and seasonal coverage are represented in Table 3.

Table 3. Monitoring period and seasonal coverage.

Item	Description
Monitoring period	6 September 2024–30 August 2025
Seasons covered	Autumn (Sep–Nov), Winter (Dec–Feb), Spring (Mar–May), Summer (Jun–Aug)
Monitoring days	6 Sep 2024; 24 Sep 2024; 24 Oct 2024; 9 Nov 2024; 17 Dec 2024; 13 Feb 2025; 13 Mar 2025; 29 May 2025; 30 Jun 2025; 30 Aug 2025

CH₄ and N₂O were recorded at 1-s intervals producing up to 86,400 measurements per stage per day. Approximately 90% of all measurements were retained after quality control and used in flux calculation. Measurements were grouped into morning (06:00 to 12:00), afternoon (12:00 to 18:00), evening (18:00 to 00:00), and night (00:00 to 06:00) for seasonal flux aggregation. Negative values were treated as zeros while occasional gaps due to operational interruptions were treated as missing data and excluded from averaging.

Environmental variables, including temperature and humidity, were obtained from the Korea Meteorological Administration (KMA) for each monitoring day and were used to assess the influence of weather conditions on GHG flux patterns.

2.3. Data Analysis and Emission Flux Calculation

2.3.1. CH₄ and N₂O Analysis

The collected GHG emission data were analyzed to assess seasonal and treatment stage variations in CH₄ and N₂O fluxes. Seasonal classification was based on local meteorological records, and data were grouped into spring, summer, fall and winter to study environmental effects on emission trends. To characterize diurnal behavior, all hourly fluxes were assigned to four 6 h periods including morning (06:00 to 12:00), afternoon (12:00 to 18:00), evening (18:00 to 00:00), and night (00:00 to 06:00). For treatment-stage characterization, emission data were categorized according to pre-treatment (raw waste storage), primary treatment (anaerobic tank I and aerobic I), and secondary treatment (anaerobic II and aerobic II) stages.

Descriptive statistics (mean, standard deviation, coefficient of variation) were used to characterize greenhouse gas. The differences among groups were evaluated using one-way ANOVA, followed by Tukey’s HSD multiple comparison test (p < 0.05) where appropriate. To quantify the influence of environmental factors, linear regression was applied with temperature and relative humidity as covariates based on the model as shown in Equation (1). All statistical analyses and graphical visualizations were performed using OriginPro 2025b SR1 (OriginLab Corporation, Northampton, MA, USA).

$$F_i={\beta }_0+{\beta }_1{T}_i+{\beta }_{2}R{H}_i+{\epsilon }_i$$

(1)

2.3.2. CH₄ and N₂O Flux Calculation

Flux rates were calculated using standard gas flux equations, integrating measured gas concentrations, air flow rates, and environmental parameters (Equation (2)).

$$F_{GHG}={\displaystyle \frac{\left(C_o-C_i\right)\times p_{air}\times \dot{m}}{T\times R\times A\times 1000}}$$

(2)

where $F_{GHG}$ is the greenhouse house gas flux $\left(\mathrm{m}\mathrm{g} \mathrm{s} {\mathrm{m}}^{-2}\right)$, $C_o$ and $C_i$ are the outlet and inlet greenhouse gas concentrations (${\mathrm{C}\mathrm{H}}_4:\mathrm{p}\mathrm{p}\mathrm{m},{\mathrm{N}}_2\mathrm{O}:\mathrm{p}\mathrm{p}\mathrm{b})$, $p_{air}$ is the atmospheric air pressure (pa), $\dot{m}$ is the air flow rate through the chamber system (${\mathrm{m}}^3 {\mathrm{s}}^{-1}$), $T$ is the ambient air temperature (K), $R$ is the molar gas constant ($\mathrm{J} {\mathrm{m}\mathrm{o}\mathrm{l}}^{-1} {\mathrm{K}}^{-1}$) and A is the effective area of the chamber (${\mathrm{m}}^2$).

2.3.3. Emission Factors Calculation

Measured fluxes for each treatment stage were aggregated across the four daily measurement periods and seasons. For each stage $s$ and season $k$ the daily representative flux was obtained as the Mean of the four period-level fluxes [9,22]. The daily emissions were calculated as:

$$E_{s,k,day}=F_{s,k,day}\times A_s\times \mathrm{86,400}\times {10}^{-6},$$

where $F_{s,k,day}$ is the daily representative flux, $A_s$ is the stage emitting surface area (${\mathrm{m}}^2$), 86,400 converts seconds to days and ${10}^{-6}$ converts $\mathrm{m}\mathrm{g}$ to $\mathrm{k}\mathrm{g}$. Seasonal total emissions were obtained by multiplying $E_{s,k,day}$ by the number of days in each season ($E_{s,k,season}$). The annual emissions per stage were then computed as follows.

$$E_{s,year}=\sum _k{E}_{s,k,season}.$$

Overall treatment plant annual emissions ($E_{total,year}$) were the sum of all stage totals. Per animal emission factors were calculated by dividing $E_{total,year}$ by the average number of pigs on the farm. $\mathrm{C}{\mathrm{O}}_2$-equivalent emission factors were derived using IPCC AR6 GWP₁₀₀ values (CH₄ = 27.2, N₂O = 273) [9].

3. Results and Discussion

3.1. Diurnal GHG Flux Characteristics

Diurnal CH₄ and N₂O fluxes showed clear 6-h interval patterns across the monitoring period (Figure 5). CH₄ fluxes were highest during the evening and night (7.2 and 7.0 $\mathrm{m}\mathrm{g} {\mathrm{s}}^{-1} {\mathrm{m}}^{-2}$ respectively), with secondary peak in the morning hours (5.8 $\mathrm{m}\mathrm{g} {\mathrm{s}}^{-1} {\mathrm{m}}^{-2}$). The lowest mean Flux occurred during the afternoon (3.2 $\mathrm{m}\mathrm{g} {\mathrm{s}}^{-1} {\mathrm{m}}^{-2}$). CH₄ variability remained comparatively narrow, with coefficient of variation (CV) ranging from 1.7 at night to 3.4 in the morning.

Figure 5. (Top) CH₄ and (Bottom) N₂O diurnal flux distributions for Morning, Afternoon, Evening, and Night periods.

N₂O fluxes exhibited far greater diurnal variability. The highest mean Flux occurred during morning hours (3174.6 $\mathrm{m}\mathrm{g} {\mathrm{s}}^{-1} {\mathrm{m}}^{-2}$), followed by a secondary night peak (2056.5 $\mathrm{m}\mathrm{g} {\mathrm{s}}^{-1} {\mathrm{m}}^{-2}$). Afternoon emissions were substantially lower, being 7.2- and 4.6-fold lower than morning and evening Fluxes, respectively. Table 4 summarizes the of diurnal CH₄ and N₂O flux statistics. The high CV values (2.6–4.8) underline the instability of N₂O production and release, which is strongly influenced by rapid operational and biochemical transitions such as changes in oxygen availability and nitrification–denitrification cycling. These findings are consistent with previous studies showing that N₂O emissions from livestock wastewater systems often occur in short-lived bursts linked to intermittent aeration and redox shifts [23,24].

Table 4. Summary of diurnal CH₄ and N₂O flux statistics from pig wastewater treatment system. Values represent the Mean, standard deviation (SD), standard error (SE), coefficient of variation (CV), and maximum observed flux (n =35 per period).

Period	GHG	Mean	SD	SE of Mean	CV	Max
Morning	CH4	5.8	19.8	3.3	3.4	116.8
	N2O	3174.6	15,199.2	2569.1	4.8	90,162.9
Afternoon	CH4	3.2	7.5	1.3	2.3	36.5
	N2O	439.9	1501.4	253.8	3.4	8671.8
Evening	CH4	7.0	14.1	2.4	2.0	79.1
	N2O	543.6	1434.7	242.5	2.6	7900.4
Night	CH4	7.2	12.4	2.1	1.7	60.5
	N2O	2056.5	8650.5	1462.2	4.2	49,913.1

3.2. Treatment Stage Specific GHG Flux Characteristics

Treatment-stage flux distributions are shown in Figure 6 and summarized in Table 5. CH₄ emissions varied markedly across the system, with the highest mean flux observed in Aeration Tank II (27.8 mg m⁻² s⁻¹). Short-term CH₄ peaks (up to 119.8 mg m⁻² s⁻¹) coincided with strong turbulence events, which likely released CH₄ accumulated within sludge flocs. Such aeration-driven stripping has been reported in both municipal and livestock wastewater systems [22,25,26,27]). Lower CH₄ fluxes in Anaerobic Tank II and the raw-water tank (2.8–3.7 mg m⁻² s⁻¹) align with reduced substrate availability, higher redox potential, and possible ammonia inhibition, conditions known to suppress methanogenesis [28,29,30,31].

Figure 6. Treatment stage variability in CH₄ and N₂O fluxes measured across the swine wastewater process.

Table 5. Summary of treatment stage specific CH₄ and N₂O flux statistics from pig wastewater treatment system. Values represent the Mean, standard deviation (SD), standard error (SE), coefficient of variation (CV), and maximum observed flux (n = 35 per stage).

Period	GHG	Mean	SD	SE of Mean	CV	Max
Aeration tank I	CH4	4.3	8.3	1.6	1.9	36.5
	N2O	2108.4	3297.3	623.1	1.6	13,803.1
Aeration tank II	CH4	14.5	27.8	5.3	1.9	116.8
	N2O	202.9	323.2	61.1	1.6	1039.3
Anaerobic tank II	CH4	2.0	2.8	0.5	1.4	10.8
	N2O	279.5	666.0	125.9	2.4	3522.4
Anaerobic Tank I	CH4	5.5	7.2	1.4	1.3	20.8
	N2O	5163.9	19,129.6	3615.2	3.7	90,162.9
Raw water storage tank	CH4	2.7	3.7	0.7	1.4	13.3
	N2O	13.5	39.9	7.5	2.9	181.8

N₂O emissions were highly stage dependent. Anaerobic Tank I produced the highest mean and peak fluxes (5163.9 mg m⁻² s⁻¹ and 90,162.9 mg m⁻² s⁻¹, respectively), reflecting unstable nitrification–denitrification processes driven by oxygen fluctuations, nitrite accumulation, and low C/N ratios. Similar extreme N₂O peaks have been observed in pig-slurry and intermittently aerated reactors, where up to 18% of influent nitrogen may be emitted as N₂O [32,33]. The high CV in Anaerobic Tank I (3.7) reinforces the episodic nature of N₂O production, consistent with wastewater studies showing large temporal variability in N₂O yields [22,34].

By contrast, Aeration Tank II and the raw-water tank exhibited minimal N₂O emissions (<300 $\mathrm{m}\mathrm{g} {\mathrm{s}}^{-1} {\mathrm{m}}^{-2}$), likely due to more stable redox environments where nitrite does not accumulate and denitrification proceeds fully to N₂ [24,35]. These contrasting emission behaviors indicate that CH₄ and N₂O mitigation strategies must target different stages and processes. CH₄ mitigation should focus on mixing and turbulence control, whereas N₂O reduction requires improving biochemical stability through controlled aeration, adequate carbon supply, and prevention of nitrite accumulation.

3.3. Seasonal GHG Flux Characteristics

Seasonal CH₄ and N₂O flux patterns are shown in Figure 7 and Table 6. CH₄ fluxes exhibited moderate but consistent seasonal variability, with the lowest values in summer (3.2 $\mathrm{m}\mathrm{g} {\mathrm{s}}^{-1} {\mathrm{m}}^{-2}$) and the highest in late fall and winter (7.0 to 7.2 $\mathrm{m}\mathrm{g} {\mathrm{s}}^{-1} {\mathrm{m}}^{-2}$). These patterns reflect temperature-driven changes in methanogenic and methanotrophic activity, as well as differences in gas–liquid transfer efficiency under cooler conditions [36]. Persistent CH₄ generation in winter likely arises from thermally buffered sludge layers [37].

Figure 7. Seasonal variations in CH₄ (top) and N₂O (bottom) fluxes across the monitoring period.

Seasonal N₂O fluxes were again dominated by Anaerobic Tank I (mean = 5163.9 $\mathrm{m}\mathrm{g} {\mathrm{s}}^{-1} {\mathrm{m}}^{-2}$, max = 90,162.9 $\mathrm{m}\mathrm{g} {\mathrm{s}}^{-1} {\mathrm{m}}^{-2}$), consistent with nitrogen transformation instability rather than climatic influence. Aeration Tank II and the raw-water tank continued to exhibit low fluxes (<300 $\mathrm{m}\mathrm{g} {\mathrm{s}}^{-1} {\mathrm{m}}^{-2}$). These results reaffirm that N₂O emissions in this system are governed primarily by operational and biochemical factors rather than seasonal temperature or humidity conditions.

Table 6. Summary of seasonal CH₄ and N₂O flux statistics from the swine wastewater treatment system. Values represent the Mean, standard deviation (SD), standard error (SE), coefficient of variation (CV), and maximum observed flux for each 6-h monitoring interval (n = 35 per period).

Seasons	GHG	Mean	SD	SE	CV	Max
Spring	CH4	5.8	19.8	3.3	3.4	116.8
	N2O	3174.6	15,199.2	2569.1	4.8	90,162.9
Summer	CH4	3.2	7.5	1.3	2.3	36.5
	N2O	439.9	1501.4	253.8	3.4	8671.8
Fall	CH4	7.0	14.1	2.4	2.0	79.1
	N2O	543.6	1434.7	242.5	2.6	7900.4
Winter	CH4	7.2	12.4	2.1	1.7	60.5
	N2O	2056.5	8650.5	1462.2	4.2	49,913.1

Table 6. Summary of seasonal CH₄ and N₂O flux statistics from the swine wastewater treatment system. Values represent the Mean, standard deviation (SD), standard error (SE), coefficient of variation (CV), and maximum observed flux for each 6-h monitoring interval (n = 35 per period).

Seasons	GHG	Mean	SD	SE	CV	Max
Spring	CH4	5.8	19.8	3.3	3.4	116.8
	N2O	3174.6	15,199.2	2569.1	4.8	90,162.9
Summer	CH4	3.2	7.5	1.3	2.3	36.5
	N2O	439.9	1501.4	253.8	3.4	8671.8
Fall	CH4	7.0	14.1	2.4	2.0	79.1
	N2O	543.6	1434.7	242.5	2.6	7900.4
Winter	CH4	7.2	12.4	2.1	1.7	60.5
	N2O	2056.5	8650.5	1462.2	4.2	49,913.1

3.4. Effect of Environmental Factors on GHG Emission Fluxes

The seasonal relationships between GHG fluxes and environmental variables are shown in Figure 8. CH₄ fluxes exhibited a very strong positive correlation with temperature (r = 0.97; R² = 0.95), confirming that methanogenesis intensifies under warmer conditions. Relative humidity (RH) showed a moderate negative correlation with CH₄ flux (r = −0.63; R² = 0.39), indicating that drier, more ventilated conditions enhance gas transfer and promote CH₄ release. These patterns are consistent with previous studies showing strong microclimatic control over CH₄ emissions from stored and aerated pig manure [36,37,38].

Figure 8. Seasonal scale linear regressions between GHG flux and temperature (○), and relative humidity (◇). Each color represents a seasonal mean: Spring, Summer, Fall, and Winter. Each point represents a seasonal mean (n = 4: Spring, Summer, Fall, Winter). Regression equations and R² values are shown for each relationship.

For N₂O, the seasonal relationship with temperature was weak (r = 0.17; R² = 0.03), suggesting that temperature plays only a minor role at the seasonal scale. In contrast, RH displayed a strong positive correlation with N₂O flux (r = 0.75; R² = 0.56), implying greater N₂O accumulation under humid, oxygen-limited conditions that favor incomplete denitrification. This finding aligns with studies showing that N₂O production in wastewater systems is driven primarily by internal biochemical dynamics rather than external temperature changes [34,35,39].

3.5. Greenhouse Gas Emission Factors

Emission factor calculation yielded annual emissions of 1.01 × 10³ kg CH₄ yr⁻¹ and 4.07 × 10⁵ kg N₂O yr⁻¹ for the wastewater treatment system. When normalized by the average herd population of 9000 pigs, the corresponding emission factors were 0.11 kg CH₄/ head/year and 45.2 kg N₂O head⁻¹ yr⁻¹. Expressed in CO₂-equivalent units using IPCC AR6 100-year global warming potentials (CH₄ = 27.2; N₂O = 273), the resulting emission factors were 3.1 kg CO₂-eq head⁻¹ yr⁻¹ for CH₄ and 1.23 × 10⁴ kg CO₂-eq head⁻¹ yr⁻¹ for N₂O.

Stage-level emissions showed different spatial patterns for CH₄ and N₂O within the treatment system (Table 7). The largest CH₄ emissions originated from Aeration Tank I, Aeration Tank II, and Anaerobic Tank I. In contrast, N₂O emissions were highly concentrated, as Anaerobic Tank I and Aeration Tank I collectively accounted for more than 90 percent of the total annual N₂O released. This concentration of N₂O emissions indicates that a small portion of the treatment process is responsible for most of the system’s greenhouse-gas output and reflects unstable nitrogen transformation conditions in those units.

Table 7. Annual CH₄ and N₂O emissions by treatment stage, percentage contributions, and system-level emission factors.

Treatment Stage	CH4 Emission (kg yr−1)	CH4 Share (%)	N2O Emission (kg yr−1)	N2O Share (%)
Aeration Tank I	252.8	25.0	115,434.3	28.3
Aeration Tank II	315.3	31.2	4258.6	1.0
Anaerobic Tank II	42.0	4.1	4906.8	1.2
Anaerobic Tank I	316.1	31.3	281,934.7	69.4
Raw Water Storage Tank	87.7	8.7	426.7	0.1
Total Annual Emissions	1014.0	100	406,961.0	100
Per-Head EF (9000 pigs)	0.11 kg CH4 head−1 yr−1	–	45.2 kg N2O head−1 yr−1	–
CO2-eq EF (IPCC AR6 GWP)	3.1 kg CO2-eq head−1 yr−1	–	1.23 × 104 kg CO2-eq head−1 yr−1	–

The CH₄ emission factor obtained in this study (0.11 kg CH₄ head⁻¹ yr⁻¹) lies at the lower edge of values reported for swine wastewater treatment systems, which typically range from 0.1–1.0 kg CH₄ head⁻¹ yr⁻¹ in full-scale biological treatment processes [22]. This consistency reflects the predominantly aerobic conditions in the treatment train, which suppress methanogenesis. In contrast, the N₂O emission factor measured here (45.2 kg N₂O head⁻¹ yr⁻¹) is several orders of magnitude higher than values reported for optimized nitrogen removal systems, where per-head N₂O emissions generally fall between 0.001 and 0.01 kg N₂O head⁻¹ yr⁻¹, and only 0.5–3% of influent nitrogen is typically emitted as N₂O-N. These elevated emissions align with the exceptionally high N₂O fluxes observed in Anaerobic Tank I and Aeration Tank I. Overall, the system’s greenhouse-gas footprint is dominated by these localized N₂O hotspots, indicating that focused improvements in aeration control, carbon availability, and nitrogen transformation processes in these units could substantially reduce total emissions.

4. Conclusions

This study quantified CH₄ and N₂O emissions from a full-scale piggery wastewater treatment system and developed treatment stage and time-resolved emission factors. CH₄ emissions were strongly influenced by temperature and gas transfer dynamics, with higher fluxes during warm periods and in aerated units, particularly during morning and night when mixing intensity and thermal gradients favored release. In contrast, N₂O emissions were dominated by biochemical instability in Anaerobic Tank I, where oxygen fluctuations, nitrite accumulation, and incomplete denitrification produced extreme and irregular peaks.

Annual emission factors were 0.11 kg CH₄ head⁻¹ yr⁻¹ and 45.2 kg N₂O head⁻¹ yr⁻¹, equivalent to 3.1 and $1.23\times {10}^{-4}$ kg CO₂-eq head⁻¹ yr⁻¹, respectively. More than 90% of N₂O emissions originated from a single treatment stage, underscoring the disproportionate influence of nitrification–denitrification instability on system-wide greenhouse-gas release.

These findings indicate that mitigation must be focused on the distinct mechanisms driving each gas. CH₄ reductions should focus on minimizing the physical release of accumulated gas, for example, by moderating aeration-induced turbulence, improving mixing uniformity to limit anaerobic microsites within aerobic reactors, and applying surface covers or gas-collection hoods in areas prone to vigorous stripping. In contrast, N₂O mitigation requires stabilization of the nitrogen transformation pathway. This involves maintaining consistent dissolved oxygen conditions to prevent abrupt redox oscillations, ensuring adequate carbon availability for complete denitrification, and employing real-time monitoring of oxygen and intermediate nitrogen species to avoid nitrite accumulation. Improving the operational stability of Anaerobic Tank I, including reducing unintended oxygen intrusion and homogenizing mixing, is likely to yield substantial reductions given its outsized contribution to total N₂O emissions.

The emission factors developed in this work enhance the accuracy of livestock greenhouse-gas inventories and offer process-specific insights that can inform low-emission wastewater treatment design. Future research will incorporate wastewater chemical characteristics and microbial community analysis to strengthen mechanistic understanding, and on developing dynamic emission factor models that integrate operational data and climatic variability to support climate-smart livestock facility management.