A comparison of CH₄, N₂O and CO₂ emissions from three different cover types in a municipal solid waste landfill

ABSTRACT

High-density polyethylene (HDPE) membranes are commonly used as a cover component in sanitary landfills, although only limited evaluations of its effect on greenhouse gas (GHG) emissions have been completed. In this study, field GHG emission were investigated at the Dongbu landfill, using three different cover systems: HDPE covering; no covering, on the working face; and a novel material-Oreezyme Waste Cover (OWC) material as a trial material. Results showed that the HDPE membrane achieved a high CH₄ retention, 99.8% (CH₄ mean flux of 12 mg C m^-2 h^-1) compared with the air-permeable OWC surface (CH4 mean flux of 5933 mg C m^-2 h^-1) of the same landfill age. Fresh waste at the working face emitted a large fraction of N₂O, with average fluxes of 10 mg N m^-2 h^-2, while N₂O emissions were small at both the HDPE and the OWC sections. At the OWC section, CH₄ emissions were elevated under high air temperatures but decreased as landfill age increased. N₂O emissions from the working face had a significant negative correlation with air temperature, with peak values in winter. A massive presence of CO₂ was observed at both the working face and the OWC sections. Most importantly, the annual GHG emissions were 4.9 Gg yr^-1 in CO₂ equivalents for the landfill site, of which the OWC-covered section contributed the most CH₄ (41.9%), while the working face contributed the most N₂O (97.2%). HDPE membrane is therefore, a recommended cover material for GHG control.

Implications: Monitoring of GHG emissions at three different cover types in a municipal solid waste landfill during a 1-year period showed that the working face was a hotspot of N₂O, which should draw attention. High CH₄ fluxes occurred on the permeable surface covering a 1- to 2-year-old landfill. In contrast, the high-density polyethylene (HDPE) membrane achieved high CH₄ retention, and therefore is a recommended cover material for GHG control.

Introduction

The waste sector accounts for approximately 5% of greenhouse gas (GHG) emissions in the global anthropogenic GHG budget (Bogner et al., 2007), and in China the total GHG amount during the processes of solid waste disposal and wastewater treatment has been estimated at 111.8 Tg CO₂-eq in 2005, accounting for 1.5% of China’s total emission (National Development and Reform Committee, 2013). Landfilling is expected to continue to be the conclusive management option within a typically adopted and regulated waste management hierarchy in the coming decades, throughout the world. In China the proportion of municipal solid waste (MSW) sent to landfills exceeds 75% (The Central People’s Government of the People’s Republic of China, 2012). This waste can release a complex mixture of gases (mainly CH₄ and CO₂) that are produced by the microbial decomposition of waste containing high levels of nitrogen and carbon (Lou and Nair, 2009), and as a result, landfills are considered to be significant sources of GHG emissions. Manfredi et al. (2009) reported direct GHG emissions of up to 300 kg CO₂-eq ton⁻¹ from mixed waste. CH₄, having a greenhouse warming potential (GWP) 25 times that of CO₂, is a major component of landfill gas (LFG), with CH₄ production in the range of 0.058~0.14 m³ kg⁻¹ waste reported in previous studies (Bogner and Spokas, 1993; Zheng et al., 2009), accounting for 2204 Gg of the total GHG emissions in China (National Development and Reform Committee, 2013). For N₂O, the amount was rather small compared to the CH₄, but it should be of more concern (Bogner et al., 2011; Harborth et al., 2013), as N₂O has a GWP 298 times that of CO₂ over a 100-year horizon. The observed N₂O fluxes for MSW landfills have been 1–2 orders of magnitude higher than those for agriculture and forest soils (Rinne et al., 2005).

The specific management patterns applied to landfills, such as capping materials or leachate irrigation, could be of prime importance in reducing GHG emissions. In a conventional landfill, soils are usually used as daily cover, intermediate cover, and even final cover. The extensive studies that have been done on GHG release from soils have shown that CH₄ fluxes varied temporally and spatially (Raco et al., 2010; Monster et al., 2015), while the cover soils at some closed landfills can become CH₄ sinks (Boeckx et al., 1996). Landfill gas (LFG) that escapes through the soil layer is the result of complex biochemical processes, including methane oxidation, nitrification, and denitrification (Zhang et al., 2008; Zhang et al., 2009). Therefore, both environmental factors and intrinsic soil properties exert interrelated effects on the production and reduction of LFG emitted from the landfill surface (Zhang et al., 2013). Other capping materials, such as pure sewage sludge, compost, and mining waste, have been used at some landfills (Börjesson and Svensson, 1997a; Sadasivam and Reddy 2014). More recently, to avoid adding volume, high-density polyethylene (HDPE) membranes have been adopted to replace conventional soil cover. Studies relevant to the effect of HDPE membranes on GHG emissions and the efficiency of the LFG recovery system are occurring (Capaccioni et al., 2011). In addition, active landfills with no cover (working face) cannot be ignored as a potential plane-shaped source of GHG (Capaccioni et al., 2011).

In order to strengthen our understanding of the effect of capping materials on GHG emissions and the interaction of multiple factors (such as landfill age, season, temperature, moisture content), it is essential to conduct field measurements of GHG emissions. In this study, GHG monitoring of three different cover types was performed using the Dongbu landfill as a case study. It was an effort to make a reliable estimate of annual cumulative GHG emissions, as well as to provide fundamental data for GHG inventory protocols and reduction potential based on long-term field observations.

Materials and methods

Sampling site

Dongbu sanitary landfill, situated in the Xiang’an District of Xiamen, Fujian Province, is an active waste treatment center. The designed waste loading of the entire landfill is 20.06 million tons with a 30-year expected operating period. The Dongbu landfill was commissioned in March 2009, and it currently has a storage capacity of 7.29 million tons. MSW is deposited daily, with an average load of 2100 tons.

The Dongbu landfill is a relatively new landfill. Three sections within the landfill with different types of engineered covers were selected for GHG emissions monitoring (Figure 1). Two were covered with HDPE and Oreezyme Waste Cover (OWC) material, respectively, and the third was the uncovered daily working face. The landfill area in our study occupied an area of 258,760 m², consisting of 227,565 m² of HDPE membrane capping, a 17,104-m² working face, a prereserved 1000 m² of OWC material capping as a trial, and the remaining area taken up by roadways. The whole landfill was installed with vertical wells for LFG extraction. Although an LFG recovery system is currently being developed to convert LFG to electricity in situ, this system is not yet in place, and consequently no LFG flow rate data can currently be obtained.

Figure 1. The panorama of Dongbu case-study landfill with (a) representative views for three sampling sites, (b) HDPE, (c) OWC, and (d) working face.

HDPE geomembranes are commonly used artificial covers for sanitary landfills in China, as both an intermediate and a final cover. The vast majority of the area was covered by HDPE without preplacement of clay or soil on the deposited waste surface, in order to increase the effective landfill volume. HDPE has the useful properties of being waterproof and impermeable to gas, in addition to its benefits of reducing the volume of the fresh leachate generated. The HDPE was welded to prevent potential gas leakage. The retained LFG would be collected by vertical wells. The sampling points were set at the surface of the HDPE that capped the 1-year-old landfill section.

OWC, a novel overlay approach, has been applied to some landfills in China. It is reported as being used for the first time in our study. OWC consists of an even layer of 0.6–1 cm thickness of muddled mixtures on the deposited refuse surface, and is created by mixing a specific composite (such as straw, bark, pulp) into a muddled paste, spraying it over the waste, and letting it dry. This material could not be degraded, whereas some fissures or cracking occurred on the OWC surface due to the drying of muddled paste. Research on this novel cover material needed to be addressed in terms of its influence on the GHGs being emitted. OWC has a dual nature of reduced hydraulic conduction and gas filtration, and thus there was no swelling. OWC is expected to replace soil or geomembranes as daily and intermediate covers and was used as a trial, covering an area of about 1000 m². In this study it was placed over a portion of the same section as the HDPE, in order to compare the effects of these two different capping materials on the LFG emissions at areas of similar age of 1 year.

In order to increase the effective landfill volume, the working face, where fresh refuse is delivered and compacted, then flattened and dumped immediately, is not covered by any capping materials at the end of a working day. This third selected measuring section presented a temporarily noncovered site, and HDPE would be installed after the section was completely filled. The location of the working face is constantly dynamic with landfill operations.

The surface gas fluxes were measured for all of the three cover types with five or six parallel chambers deployed randomly and sampled simultaneously at each point. To determine the long-term emission profile of the landfill, field campaigns were conducted about every 2 weeks over a 1-year time frame, from April 2012 to April 2013, so as to conduct 21 sampling rounds. Long-term and intensive sampling rounds have supplied reliable data for GHG emissions.

Flux measurements of three GHGs

This study was conducted using the static chamber method. The chamber, made of polyvinyl chloride polymer (PVC), constituted a combination of a round base and a portable top, with a cross-sectional area of 0.13 m² and a volume of 38 L. The chamber was protected with Styrofoam on the outside to prevent solar radiation from raising the headspace temperature during the sampling period. According to various sampling sites, multiple bases have been designed for use with different cover types, and these were mostly installed prior to sampling. For the HDPE geomembrane covering, PVC bases with water-sealed rims were stuck directly to the surface. Typical insertion of bases would puncture the HDPE cover and create a large measurement deviation. Some fissures or cracking on the OWC surface should be avoided when placing the base. The base frame collar was inserted directly into the OWC layer. However, it was not possible to fix the bases permanently on the working face because of high traffic volume and frequent shifts of the working surface. Therefore, field installation of bases should be done on the working face just prior to chamber sampling (average insert depth = 8 cm).

Before each sampling, a sealed chamber was created by filling a rim around the base with water to prevent gas leakage when the top chamber was placed onto the rim, and was subsequently equipped with a pressure equilibrium tube. The enclosed headspace gas was then extracted into a 60-ml syringe with a gas-tight three-way valve assembled in the sampling port, at 5- or 10-min intervals as determined by pre-experiments, five times over a period of 20 min for the OWC-covered area and the working face, and of 40 min for the HDPE-covered area. Flux measurements were carried out at in the morning (9:00 a.m. to noon). The changes in the headspace temperature (JM 624, Jinming Co. Ltd, Tianjin, China) in the sealed chambers and the MSW moisture contents (MP 406, Zhongti Co. Ltd, Nantong, Jiangsu, China) of the working face (0–6 cm) using the portable probe were also recorded for each test period.

Every group of five sequential gas samples was analyzed within 24 hr using a gas chromatograph (GC, Agilent 7890A, Palo Alto, CA), and GHG fluxes were calculated based on the change in gas concentrations over time (dC/dt), chamber height, and temperature correction (Zhang et al., 2013).

Statistical analyses were carried out for the purpose of identifying the correlation between GHG emissions and environmental factors using bivariate regression analysis. The differences were conducted by one-way analysis of variance (ANOVA), and p < 0.05 represented a 95% confidence level.

Results and discussion

GHG emissions from the three cover types

HDPE

As Figure 2 illustrates, GHG emission fluxes were small throughout the HDPE-covered section during the sampling rounds and no obvious change appeared—a result most likely due to the low emission fluxes. Table 1 also concluded the statistical measures for target gas emission fluxes at each sampling section. Taking CH₄ as an example, only 0.13–52.7 mg C m⁻² hr⁻¹ was detected, of which the mean was 12.20 mg C m⁻² hr⁻¹. For N₂O, all the values were below 0.1 mg N m⁻² hr⁻¹, with the mean and median values of 0.021 mg N m⁻² hr⁻¹ and 0.010 mg N m⁻² hr⁻¹, respectively. Zhang et al. (2013) also observed the lowest fluxes of CH₄ and N₂O on the HDPE surface, with a 40- to 51.6-fold lower CH₄ flux compared with the other sampling sections. Moreover, CO₂ emissions were also sustained at a relatively low level in this study, ranging from 0.88 to 75.0 mg C m⁻² hr⁻¹. The CO₂ flux values could not indicate the degradation rate of organic carbon here, since HDPE membrane effectively retained LFG leading to a strong interference with CO₂ release.

Table 1. Dongbu case-study landfill: Statistical measures for GHG emission flux monitoring campaigns.

Landfill section	GHG	Number of data points	Mean	Median	Minimum	Maximum	SD	CV(%)
HDPE	CH4	73	12.20	6.35	0.13	52.66	12.66	103.5
	N2O	55	0.020	0.010	0.000	0.100	0.024	119.1
	CO2	74	17.37	14.52	0.88	74.97	13.15	75.7
OWC	CH4	91	5932.8	2437.3	3.7	31188.3	7906.0	133.3
	N2O	76	0.247	0.149	0.005	1.149	0.256	103.6
	CO2	96	5529.5	3178.4	86.6	27883.5	6330.9	114.5
Working face	CH4	100	318.4	77.3	2.7	3672.9	681.1	213.9
	N2O	109	10.42	3.33	0.01	99.67	17.43	167.3
	CO2	107	8220.7	6726.4	176.6	20494.6	6792.1	82.6

Note. GHG flux units are mg C m⁻² hr⁻¹ for CH₄ and CO₂, and mg N m⁻² hr⁻¹ for N₂O. SD: standard deviation. CV: coefficient of variation.

Figure 2. GHG emission fluxes of HDPE membrane surface for 20 sampling rounds. Bars indicate positive standard deviation. Lines indicate average values based on five or six parallel measurements at each sampling round.

OWC

The GHG emission profiles on the OWC surface during the monitoring are shown in Figure 3. The maximum CH₄ fluxes were observed at this section, in the range of 3.7–31188.3 mg C m⁻² hr⁻¹, with the mean value of 5932.8 mg C m⁻² hr⁻¹. The mean value was lifted upwards by some fluxes of greater than 10 g C m⁻² hr⁻¹. The strong CH₄ fluxes were 3–5 orders of magnitude higher than those from the HDPE-covered section with the same landfill age (p < 0.05). The waste pile covered by 15–30 cm soil with a landfill age of 1 year had a CH₄ flux of 0–21716.7 mg C m⁻² hr⁻¹ at another landfill site (Abichou et al., 2006), comparable to the OWC surface. The N₂O release at this section was still minor, in the range of 0.005–1.15 mg N m⁻² hr⁻¹. Ninety-eight percent of the monitoring data showed that the fluxes varied in the range of micrograms per square meter, and were similar to those of other landfills (Zhang et al., 2008). At this section, intense CO₂ emissions were indeed observed. CO₂ fluxes varied from 86.6 to 27,883.5 mg C m⁻² hr⁻¹ and the average value was 5529.5 mg C m⁻² hr⁻¹. The mean flux ratio of CH₄ and CO₂ was calculated at 1.07, showing a prevalence of the dominant CH₄ condition (Capaccioni et al., 2011). These results implied that the waste deposited inside the landfill column with a 1- to 2-year landfill age entered the methanogenesis stage, creating considerable CH₄ and CO₂ produced under anaerobic conditions and vented directly to the ambient atmosphere through the permeable OWC cover. Furthermore, all three kinds of measured GHGs were higher in the first 8 months than in the last 4 months. The average fluxes of CH₄ and CO₂ in the last 4 months dropped to 30% of the annual mean value of the emissions. The period of the last 4 months had a longer landfill age. This result was possibly ascribed to the decreasing trend of landfill gas generation with the increasing landfill age.

Figure 3. GHG emission fluxes of the surface of OWC capping materials. Bars indicate positive standard deviation.

Comparing the two cover materials used on the landfill sections with the same landfill age, CH₄ emissions on the HDPE section came to only 0.2% of those of the OWC section (p < 0.05), suggesting that the vast majority of CH₄ generated under anaerobic conditions could be effectively retained by HDPE and collected by the LFG recovery system, resulting in higher recovery efficiency. For example, when in another study an HDPE cover system was increased from 25% to 32% of the area of the Bellolampo sanitary landfill, the recovery efficiency of the LFG system was significantly enhanced, from 79.5% to 88.4% (Trapani et al., 2013). However, the air-permeable OWC cover seemed to emit considerable CH₄ and CO₂ from the waste pile entering the methanogenesis stage. On the other hand, HDPE was a low-permeability material toward water and gas, and the OWC section was adjacent to the HDPE section. This led to a preferable way of the potential lateral migration and concentration of LFG from the extended HDPE section to the permeable OWC section, which was also responsible for the phenomenon of higher fluxes of the OWC section compared to the HDPE section to an extent. Our results were in line with the findings by Capaccioni et al. (2011). Actually, OWC material is expected to replace soil as a daily or intermediate cover for saving space in the studied landfill. However, considering its inability to control GHG emissions, OWC capping seems an unsuitable option, while the alternative HDPE membrane appears more favorable.

Working face

The fluctuations of N₂O and CH₄ fluxes on the working face were completely different from those at the OWC section (Figure 4). Among these, CH₄ fluxes displayed a range of 2.7–3672.9 mg C m⁻² hr⁻¹, which were about 20–30 times lower than those at the OWC section (p < 0.05). The results indicated that the presence of oxygen at the early stage of a landfill works against methanogenesis. Only when the waste pile rapidly shifted to anaerobic conditions could methanogens be active and reproducible, and finally become the dominant species.

Figure 4. GHG emission fluxes of the working face for the newly dumped waste. Bars indicate positive standard deviation.

In addition, newly dumped waste produced substantial amounts of N₂O at the initial adjustment stage (Harborth et al., 2013). Our study found a rapid rise in N₂O emissions soon after landfilling. Significant N₂O emissions were observed in the range of 0.01–99.67 mg N m⁻² hr⁻¹ (CV = 167%) on the working face, especially up to more than 60 mg N m⁻² hr⁻¹ at some particular points. The mean and median N₂O fluxes were 10.42 and 3.33 mg N m⁻² hr⁻¹, respectively. The high N₂O emissions from the working face can probably be ascribed to elevated contents of NH₄⁺-N and NO₃^–N in the fresh waste (Sormunen et al., 2008). First, NH₄⁺ was identified as the substrate for nitrification to produce NO₃⁻, and these original and generated NO₃⁻ then underwent further denitrification in the anaerobic niches of the compacted waste. Both nitrification and denitrification have been reported to be the pathway for N₂O production in landfills (Mandernack et al., 2000; Harborth et al., 2013). The N₂O emission peak occurred in the early stage of MSW dumping (Sormunen et al., 2008; Bogner et al., 2011). Harborth et al. (2013) also observed that N₂O fluxes on the working face were in the range of 0–39 mg N m⁻² hr⁻¹, with a high value of 428 mg N m⁻² hr⁻¹. In our study, N₂O emissions spanned four orders of magnitude during the one-year monitoring period, and fluxes at the milligram level accounted for a large fraction of 66% of the total flux data. Hence, N₂O emitted from the working face should not be ignored, since N₂O has a greenhouse warming potential 298 times that of CO₂. Furthermore, CO₂ fluxes at this section were comparable to the OWC-covered 1- to 2-year-old landfill, indicating that organic carbon started to degrade as soon as the waste was dumped, and this high degradation rate could continue for a relatively long time.

Correlation analysis

The capping materials significantly interfered with the effect of environmental factors on GHG emissions. Herein we discuss the relevance between the three target gases and environmental factors on the working face (Table 2). Here, CO₂ fluxes were correlated with CH₄ (p < 0.05), and also had a significant correlation with the N₂O (p < 0.01). Denitrification was more active than methanogenesis in the presence of oxygen, with organic carbon as the electron donor during the reduction of nitrogen species simultaneously coupled with a considerable amount of CO₂ production.

Table 2. Correlation analysis for GHG emissions and relevant evironmental factors on the working face section.

	CH4	N2O	CO2	Moisture	Air temperature
CH4	1.00	–0.214	0.273*	–0.308**	0.446**
N2O		1.00	0.345**	0.575**	–0.331**
CO2			1.00	0.108	0.332**
Moisture				1.00	–0.360**
Air temperature					1.00

Note. N = 126. Asterisk indicates correlation is significant at the 0.05 level; double asterisk, correlation is significant at the 0.01 level.

Moreover, GHG emissions were affected by environmental factors, such as air temperature and moisture content (Table 2). High temperatures in the waste pile were beneficial to the degradation of organic carbon (Zhang et al., 2013) and to CH₄ generation by methanogens (Weitz et al., 2001); hence air temperature was significantly positively correlated with CH₄ and CO₂ emissions (p < 0.01). However, when temperature rises, the activity of N₂O reductase involved in denitrification increases and N₂O has a higher potential to convert to N₂, leading to a decline in N₂O emissions (Holtan-Hartwig et al., 2002). Hence, correlation analysis showed a negative relationship between air temperature and N₂O fluxes (p < 0.05). Air temperature and moisture are strongly negatively correlated (p < 0.01), and thus the temperature effect can also be a moisture effect. Waste moisture content had a significant positive correlation with N₂O emissions (p < 0.01). Extrapolation shows that water contained in newly dumped waste created a pattern of alternating wet and dry conditions that led to alternating nitrification and denitrification, promoting N₂O production (Weitz et al., 2001). Generally, CH₄ oxidation is an aerobic process brought about by methanotrophic bacteria, so that a modest improvement in soil moisture content resulting in the reduction of O₂ content would cause an increase in CH₄ fluxes on the surface of the waste pile (Barlaz et al., 2004), showing a positive relationship between CH₄ flux and soil moisture. CH₄ emissions here were negatively correlated with moisture content (p < 0.05), which indicated that an increase in moisture content might decrease the ability of CH₄ to be migrated from the waste pile to the ambient atmosphere. Gas migration and moisture must be considered under simultaneous reaction of aerobic and anaerobic type, depending on waste depth.

Seasonal variations of GHG emissions

Several observations can be made about seasonal changes in GHG emissions from the three different cover types (Figure 5). First, since almost no gas escaped from the waste covered with the HDPE membrane, no seasonal trends of GHG emissions were observed there. At the OWC section, the seasonal variations here followed the order spring ≈ summer > autumn > winter. The seasonal fluctuations in temperature and water content probably influenced the methanogens activities inside the waste pile. Similarly, CO₂ emissions also exhibited a trend of higher fluxes in spring, summer, and autumn, with lower values in winter. Contrary to our results, another study observed lower CH₄ flux at a soil-covered landfill in the summer (Börjesson and Svensson, 1997b), which can be ascribed to an increase in CH₄ oxidation in landfill soil with increasing temperature (Henneberger et al., 2015; Zhang et al., 2013). In contrast to other previous research, OWC materials with 0.6–1 cm thickness seemed to have no ability to oxidize CH₄, and consequently bacterial metabolism under high-temperature conditions was so active that it accelerated a massive CH₄ escape. On the other hand, when we extended the monitoring period in our study, LFG generation decreased with an increase in landfill age (Bella et al., 2011), due to the depletion of organic matter; this in turn caused a reduction in GHG emissions.

Figure 5. Seasonal variations of N₂O, CH₄, and CO₂ fluxes at the three landfill sites. The given data represent the mean values of the corresponding fluxes measured in spring (April–May 2012; March–April 2013), summer (June–August 2012), autumn (September–November 2012), and winter (December–February 2012–2013), respectively. Bars indicate positive standard deviation.

The working face was identified as a nitrification–denitrification hotspot. The N₂O fluxes in winter were obviously higher than those of the other three seasons. The N₂O reductase was more sensitive at low temperature than other denitrifying reductase, leading to an inhibition of the conversion of N₂O to N₂ (Holtan-Hartwig et al., 2002). Relatively strong N₂O emissions were also detected in winter in arable soils (Flessa et al., 1995). Moreover, the fact that CO₂ emissions were more prevalent in summer but roughly equal in the other three seasons, which implied that the biodegradation of organic matter was activated under higher temperatures.

Evaluation of cumulative annual GHG emissions in the case-study landfill

We assumed the average measured fluxes of the three target gases were representative of the release characteristics for each specific section despite seasonal and temporal variations. The cumulative annual GHG emissions were computed by multiplying by the corresponding areas of the study sections, as summarized in Table 3.

Table 3. Target gas emissions characteristics of the specific sites and the whole landfill.

Sources	Annual emissions (t yr^−1)			Annual emissions (GWP100 CO2-eq t yr^−1)			GWP total	Emissions from per unit area
	CH4	N2O	CO2	CH4	N2O	CO2
HDPE	32.4	0.066	127.0	811.0	19.7	127.0	830.7	17.0
OWC	69.3	0.003	177.6	1732.5	1.0	177.6	1733.5	35.5
Working face	63.6	2.453	4516.3	1590.2	731.1	4516.3	2321.3	47.5
GHG total	165.3	2.522	—	4133.7	751.8	—	4885.5
Percentage	—	—	—	84.6	15.4	—	100.0

As to the HDPE membrane, which covered 88% of the total landfill area, the cumulative emission amounts of CH₄, N₂O, and CO₂ were 32.4, 0.066, and 127.0 t yr⁻¹, respectively, which contributed the relatively low percentages of 19.6%, 2.6%, and 2.6% of the annual emissions of the three gases, respectively. Special emphasis should be given to the GHG emissions at the OWC section. The section, with a landfill age of 1–2 years, generated substantial CH₄ after entering the methanogenesis stage, and as a result, the annual CH₄ emissions of the OWC (1732.5 t yr⁻¹) were 2.1 times as high as those from the HDPE, in spite of a smaller coverage area of 1000 m², comprising 41.9% of the total amount of CH₄ emissions. Moreover, although the working face occupied only 6.6% of the overall area, it was definitely the N₂O hotspot. The N₂O emission fluxes at the working face varied at the milligram level, in contrast to the microgram level of the N₂O fluxes at both the HDPE and the OWC sections; consequently, the working face became by far the highest contributor to N₂O emissions, accounting for over 97% of the N₂O source. Currently, although vertical wells for LFG extraction have been installed, the LFG recovery system has not been completed, and the majority of captured LFG is still being vented to the atmosphere on site rather than being further utilized. Hence, no captured LFG data could be obtained for our study. The collected LFG would be converted to electricity and the flow data could be obtained once the recovery system comes into service. Meanwhile, a fraction of the targeted gases is emitted through leachate collection pipelines or treatment plants (Wang et al., 2014). Another reported case study (Capaccioni et al., 2011) showed that 83–85% recovery efficiency was achieved, according to multiyear field measurements on an LFG recovery system using HDPE as a cover. If the collected LFG were converted on site to electricity at the Dongbu landfill, we could achieve a win–win outcome of simultaneous energy recovery and GHG reduction.

The three investigated sections comprised all of the GHG sources. The annual GHG emissions expressed as greenhouse warming potential are also listed in Table 3. Since CO₂ is derived from biomass carbon, that is, with carbon neutrality, CO₂ was not contained in the inventory estimates of GHGs. The GHG amounts from the HDPE, the OWC, and the working face accounted for 17.0%, 35.5%, and 47.5%, respectively. It can be seen that adopting HDPE membrane as a covering material could be an important strategy for controlling GHGs. However, HDPE only prevents the LFG emissions from the landfills rather than reducing LFG generation or enhancing the LFG removal. If no LFG recovery system was operated, the retained LFG would escape through the vertical wells, and the HDPE cover could not efficiently control GHG emission from this point of view. The sum of the GHG emissions came to 4.89 Gg yr⁻¹; CH₄ was clearly the main GHGs in the landfill, while the N₂O emissions from the working face comprised 15.0% of the total GHG emissions. Thus, separating the landfill into several cells might result in a reduction of the working face area, further mitigating the N₂O emissions.

Conclusion

Capping materials with different characteristics exerted a remarkable effect on GHG emissions at the landfill sites. The HDPE section had a high CH₄ mitigating effect on observed emissions of 99.8% compared to the adjacent air-permeable OWC section with the same landfill age. OWC material is expected to be used as a daily or intermediate cover; therefore, OWC capping seems an unsuitable option from the view of GHG control. By contrast, the working face was the hotspot of N₂O emission. The annual GHG emissions were 4.89 Gg yr⁻¹ in CO₂ equivalents for the landfill as a whole, and CH₄ constituted the majority. In terms of the effects of the different coverings, covering most of the landfill area with HDPE membrane would create a lower risk for GHG escape. HDPE is therefore recommended as a capping material for GHG control.

Acknowledgment

The authors acknowledge the staff at the Dongbu MSW landfills in Xiamen for the survey and sample collaboration.

Funding

This research was supported financially by the “Strategic Priority Research Program—Climate Change: Carbon Budget and Relevant Issues” of the Chinese Academy of Sciences (grant XDA05020602) and the National Natural Science Foundation of China (grant 41475130).

Additional information

Funding

This research was supported financially by the “Strategic Priority Research Program—Climate Change: Carbon Budget and Relevant Issues” of the Chinese Academy of Sciences (grant XDA05020602) and the National Natural Science Foundation of China (grant 41475130).

Notes on contributors

Xiaojun Wang

Xiaojun Wang, Xiangyu Lin, Ying Xu, and Xin Ye are research assistants at the Institute of Urban Environment, Chinese Academy of Sciences, in Xiamen, China.

Mingsheng Jia

Mingsheng Jia is a master’s student at the Institute of Urban Environment, Chinese Academy of Sciences, in Xiamen, China.

Xiangyu Lin

Xiaojun Wang, Xiangyu Lin, Ying Xu, and Xin Ye are research assistants at the Institute of Urban Environment, Chinese Academy of Sciences, in Xiamen, China.

Ying Xu

Xiaojun Wang, Xiangyu Lin, Ying Xu, and Xin Ye are research assistants at the Institute of Urban Environment, Chinese Academy of Sciences, in Xiamen, China.

Xin Ye

Xiaojun Wang, Xiangyu Lin, Ying Xu, and Xin Ye are research assistants at the Institute of Urban Environment, Chinese Academy of Sciences, in Xiamen, China.

Chih Ming Kao

Chih Ming Kao is a research professor at the Institute of Environmental Engineering, National Sun Yat-Sen University, in Kaohsiung, Taiwan.

Shaohua Chen

Shaohua Chen is a research professor at the Institute of Urban Environment, Chinese Academy of Sciences, in Xiamen, China.