Characterization of Gas Transport Properties of Compacted Solid Waste Materials

Gas transport parameters such as gas diffusivity (Dp/D0), air permeability (ka), and their dependency on void space (air-filled porosity, ε) in a waste body govern convective air and gas diffusion at solid waste dumpsites and surface emission of various gases generated by microbial processes under aerobic and anaerobic decompositions. In this study, Dp/D0(ε) and ka(ε) were measured on dumping solid waste in Japan such as incinerated bottom ash and unburnable mixed waste as well as a buried waste sample (dumped for 20 years). Sieved samples with variable adjusted moistures were compacted by a standard proctor method and used for a series of laboratory tests for measuring compressibility, saturated hydraulic conductivity, and gas transport parameters. Results showed that incinerated bottom ash and unburnable mixed waste did not give the maximum dry density and optimum moisture content. Measured compressibility and saturated hydraulic conductivity of tested samples varied widely depending on the types of materials. Based on the previously proposed Dp/D0(ε) models, the diffusion-based tortuosity (T) was analyzed and unique power functional relations were found in T(ε) and could contribute to evaluating the gas diffusion process in the waste body compacted at different moisture conditions.


Introduction
Solid waste disposal has always remained a crucial part of the waste management system. Due to the scarcity of landfills, it is essential to consider landfill facilities' optimization and how ecofriendly they are. For volume reduction, burnable waste is usually incinerated and unburnable waste shredded before dumping at landfill sites in Japan [1,2]. In contrast, in developing countries, without segregation waste is directly sent to the landfill site, consisting of heterogeneous materials [3,4]. The decomposition of the dumped solid waste produces various types of organic and inorganic gases such as methane (CH 4 ), carbon dioxide (CO 2 ), aromatic carbons, nonmethane hydrocarbons, and hydrogen sulfide (H 2 S) [5][6][7][8]. It is also well known that waste landfills dumped with organic-rich waste are a significant source of anthropogenic methane emission resulting from the microbial degradation of organic matter in a waste body [9]. The microbial degradation changes aerobic to anaerobic degradation over a specific time depending upon the landfill operation due to the reduction of void spaces [10][11][12].
To understand the gas emission from waste dumping sites and to reduce the harmful gas generation, it is essential to evaluate the gas transport process in the waste body as well as the landfill cover soil and to maintain the proper air movement in the waste body [10,13]. The gas transport processes in the waste body and emission of landfill gases are controlled by many factors such as packing of dumped waste (i.e., compaction), moisture and void space, degree of biodegradation, the thickness of landfill cover soil, and the design of dumpsite facility [14,15]. In particular, gas transport parameters of gas diffusivity (D p /D 0 : the ratio of gas diffusion coefficient, D p , to gas diffusion in free air, D 0 ) and air permeability (k a ) and their dependency on void space (air-filled porosity, ε) in the waste body govern convective air and gas diffusion at solid waste dumpsites. However, only limited information on gas diffusivity and air permeability for compacted dumped waste are available.
Therefore, this research aims to measure the gas transport parameters, D p /D 0 and k a , for compacted typical dumping solid waste such as incinerated bottom ash and unburnable mixed waste (after shredded) and to characterize their dependency of ε [= air-filled volume (cm 3 )/total volume (cm 3 )]. Besides, a buried waste sample (dumped for 20 years) taken from a waste dumping site was used as a testing material as well as literature data of a landfill cover soil [13] for evaluating the pore tortuosity characteristics based on the previously proposed D p /D 0 models.

Tested Waste Materials
In this research, three different types of waste brought to an engineered waste landfill in Saitama prefecture, Japan, were used: incinerated bottom ash from municipal solid waste and two types of unburnable mixed waste [hereafter labeled as unburnable mixed waste (1) and (2)]. Besides, buried waste was taken from a post-closure industrial waste landfill site in Saitama Prefecture, Japan. The buried waste was taken around 2 m below ground surface and was dumped for 20 years, according to the landfill record. Photos of samples are shown in Figure 1.

Physical and Chemical Properties, Waste Composition
Basic physical and chemical properties of the tested samples were determined following American Standards for Testing and Materials (ASTM) and Japan Industrial Standards (JIS) and were presented in Table 1. According to Unified Soil Classification System (USCS) textural classification, incinerated bottom ash and buried waste (dumped for 20 years) belonged to coarse grained sand, while unburnable mixed waste (1) and (2) belonged to coarse grained gravel. Size distributions of the tested samples were shown in Figure 2 together with the reported particle size distributions of a landfill cover soil, <2 mm and <35 mm, taken from the same engineered landfill in Saitama Prefecture [13]. REVIEW 2 To understand the gas emission from waste dumping sites and to reduce the harm gas generation, it is essential to evaluate the gas transport process in the waste bod well as the landfill cover soil and to maintain the proper air movement in the waste b [10,13]. The gas transport processes in the waste body and emission of landfill gases controlled by many factors such as packing of dumped waste (i.e., compaction), mois and void space, degree of biodegradation, the thickness of landfill cover soil, and the sign of dumpsite facility [14,15]. In particular, gas transport parameters of gas diffusi (Dp/D0: the ratio of gas diffusion coefficient, Dp, to gas diffusion in free air, D0) and permeability (ka) and their dependency on void space (air-filled porosity, ε) in the w body govern convective air and gas diffusion at solid waste dumpsites. However, o limited information on gas diffusivity and air permeability for compacted dumped w are available.
Therefore, this research aims to measure the gas transport parameters, Dp/D0 and for compacted typical dumping solid waste such as incinerated bottom ash and unbu ble mixed waste (after shredded) and to characterize their dependency of ε [= air-fi volume (cm 3 )/total volume (cm 3 )]. Besides, a buried waste sample (dumped for 20 ye taken from a waste dumping site was used as a testing material as well as literature d of a landfill cover soil [13] for evaluating the pore tortuosity characteristics based on previously proposed Dp/D0 models.

Tested Waste Materials
In this research, three different types of waste brought to an engineered waste lan in Saitama prefecture, Japan, were used: incinerated bottom ash from municipal s waste and two types of unburnable mixed waste [hereafter labeled as unburnable mi waste (1) and (2)]. Besides, buried waste was taken from a post-closure industrial w landfill site in Saitama Prefecture, Japan. The buried waste was taken around 2 m be ground surface and was dumped for 20 years, according to the landfill record. Photo samples are shown in Figure 1.

Physical and Chemical Properties, Waste Composition
Basic physical and chemical properties of the tested samples were determined lowing American Standards for Testing and Materials (ASTM) and Japan Indus Standards (JIS) and were presented in Table 1. According to Unified Soil Classifica System (USCS) textural classification, incinerated bottom ash and buried waste (dum for 20 years) belonged to coarse grained sand, while unburnable mixed waste (1) and belonged to coarse grained gravel. Size distributions of the tested samples were show Figure 2 together with the reported particle size distributions of a landfill cover soil mm and <35 mm, taken from the same engineered landfill in Saitama Prefecture [13].   Incinerated bottom ash and buried waste consisted mainly of the size fraction w <5 mm. Measured pH values showed that incinerated bottom ash (pH = 11.1) and bur waste (pH = 8.8) were alkaline, while unburnable mixed waste (1) and (2) (pH = 7.2, including landfill cover soil (pH = 5.6) were neutral. Measured loss on ignition (L showed the highest value (81%) for unburnable mixed waste (2) due to the presence of rubber and plastic types fraction (see, Figure 3).
After air-drying the tested samples, the waste composition for each tested sam was determined using around 2-3 kg air-dried mass, and the results are shown in Fig  3. Incinerated bottom ash and buried waste consisted mainly of residues (<5 mm) incl ing a soil-like material (see also Figure 2). Unburnable mixed waste (1) contained arou 50% of glass and ceramics; on the other hand, unburnable mixed waste (2) contained h amounts of plastic and rubber (see also Figure 1). Unburnable mixed waste (1) Unburnable mixed waste (2) Landfill cover soil < 2mm Landfill cover soil < 35mm  Incinerated bottom ash and buried waste consisted mainly of the size fraction with <5 mm. Measured pH values showed that incinerated bottom ash (pH = 11.1) and buried waste (pH = 8.8) were alkaline, while unburnable mixed waste (1) and (2) (pH = 7.2, 7.8) including landfill cover soil (pH = 5.6) were neutral. Measured loss on ignition (LOI) showed the highest value (81%) for unburnable mixed waste (2) due to the presence of the rubber and plastic types fraction (see, Figure 3).
After air-drying the tested samples, the waste composition for each tested sample was determined using around 2-3 kg air-dried mass, and the results are shown in Figure  3. Incinerated bottom ash and buried waste consisted mainly of residues (<5 mm) including a soil-like material (see also Figure 2). Unburnable mixed waste (1) contained around 50% of glass and ceramics; on the other hand, unburnable mixed waste (2) contained high amounts of plastic and rubber (see also Figure 1). Unburnable mixed waste (1) Unburnable mixed waste (2) Landfill cover soil < 2mm Landfill cover soil < 35mm Incinerated bottom ash and buried waste consisted mainly of the size fraction with <5 mm. Measured pH values showed that incinerated bottom ash (pH = 11.1) and buried waste (pH = 8.8) were alkaline, while unburnable mixed waste (1) and (2) (pH = 7.2, 7.8) including landfill cover soil (pH = 5.6) were neutral. Measured loss on ignition (LOI) showed the highest value (81%) for unburnable mixed waste (2) due to the presence of the rubber and plastic types fraction (see, Figure 3).
After air-drying the tested samples, the waste composition for each tested sample was determined using around 2-3 kg air-dried mass, and the results are shown in Figure 3. Incinerated bottom ash and buried waste consisted mainly of residues (<5 mm) including a soil-like material (see also Figure 2). Unburnable mixed waste (1) contained around 50% of glass and ceramics; on the other hand, unburnable mixed waste (2) contained high amounts of plastic and rubber (see also Figure 1).

Measurements of Compaction Property, Compressibility, Hydraulic Conductivity, and Gas Transport Parameters
A series of laboratory tests were carried out to determine compaction property, compressibility, hydraulic conductivity, and gas transport parameters. A chart of sample preparation, treatment, and subsequent measurements is shown in Figure 4. First, the samples were sieved with <2 mm and/or <9.5 mm, and adjusted moisture contents. Then, the samples were used for the compaction tests following a standard Proctor method [16] and determined the maximum dry density (ρ dmax ) and optimum moisture content (OMC). In the compaction test, the samples were packed into three layers in a mold with an internal   The compacted samples with different moisture conditions packed in a mold with an inner diameter of 15 cm and a height of 12 cm were used for measuring the gas transport parameters, D p /D 0 and k a . A gas diffusion chamber was used to measure D p /D 0 at room temperature at 20 • C, and oxygen was used as a tracer gas [17]. The diffusion chamber was first flushed with 100% N 2 , after which the upper end of the testing sample was exposed to the atmosphere. Oxygen was measured in the diffusion chamber with an oxygen electrode. Change of oxygen was measured as a function of time in the diffusion chamber, and the D p (cm 2 /s) was calculated by applying Fick's law. The gas diffusion of oxygen in free air (D 0 ) at 20 • C of 0.20 (cm 2 /s) was used in this research. The k a (µm 2 ) was measured at room temperature at 20 • C using a steady-state method [18]. A constant, small air pressure difference (to maintain laminar steady airflow without the redistribution of water inside the sample) was applied to the samples, and the airflow rate was measured. The k a was calculated based on Darcy's law using air density and viscosity. Then, the measured D p /D 0 and k a values were plotted against measured air-filled porosity (ε). It was noted that the ε values for compacted samples were calculated using dry bulk density and water content of tested samples in this study, indicating that the ε values did not represent connected open pores, but a total volume of air-filled pores in the samples. Additionally, the compaction property, compressibility, and hydraulic conductivity for unburnable mixed waste (1) and (2) were measured but, the D p /D 0 and k a values for these samples could not be determined because a gap between the compacted sample and surrounding mold caused the inaccuracy of measured values.

Compaction, Compressibility, and Hydraulic Properties
Measured compaction curves of the tested samples are shown in Figure 5. Incinerated bottom ash (Figure 5a) and unburnable mixed waste (1) (Figure 5c), and unburnable mixed waste (2) (not shown) were independent of initial moisture condition and did not show any clear peak (i.e., the maximum dry density, ρ dmax ) in the compaction curves. Similar results were reported in many kinds of literature, especially for the tested results from cohesionless solid waste materials and industrial by-products, which did not show the clear peak in the compaction curves: incinerated bottom ash [19,20], recycled concrete and clay bricks [21], glass types materials [22], and slag bottom ash [23]. On the other hand, buried waste (dumped for 20 years) (Figure 5b) as well as landfill cover soil (data not shown. See Figure 3 in [13]) showed a clear peak and gave ρ dmax and optimum moisture content (OMC) in the compaction curve.    The measured values of ρ dmax and OMC were summarized in Table 2 with some reported literature values. Generally, the ρ dmax values depending on the material properties such as specific gravity (G s ) and composition, lower G s materials gave lower ρ dmax values. This can be found typically for the unburnable mixed waste (1) rich in plastic and vinyl in this study and dumped municipal solid waste (MSW) rich in organic materials [24]. Besides, it can be found that the materials with coarser fraction showed higher ρ dmax and lower OMC compared to the materials with finer fraction (e.g., buried waste in this study and landfill cover soil [13]).
Measured values of compressibility (C c ) and saturated hydraulic conductivity (k s ) are summarized in Table 2. Among tested samples in this study, unburnable mixed waste (2) showed the highest C c (=0.23). This was probably due to the unburnable mixed waste (2), which was rich in easily compressible materials such as plastic, vinyl, and rubber (Figure 3d). On the other hand, the unburnable mixed waste (1) rich in glass and ceramics gave the lowest C c (=0.013) due to the non-compressible nature of those materials. Compared to the reported C c values of dumped and fresh MSW [24][25][26][27][28], the measured C c values of incinerated bottom ash and buried waste in this study became approximately one order smaller.
Measured k s values for unburnable mixed waste (1) and (2) were in the order of 10 −4 m/s and became higher than those for other materials including dumped and fresh MSW. This indicated clearly that the waste body consisted of an unburnable mixed waste of plastic, glasses, ceramics, which became more water permeable, contrary to other dumped industrial by-products of incinerated bottom ash (10 −6 m/s), MSW (10 −8~1 0 −4 m/s) and buried waste with high fine residue (10 −8 m/s). It was noted that k s values in the dumped waste body change widely with the stress level, density due to the overburden, and strata of dumped waste in the landfill site [27,29]. Besides, the decomposition of biodegradable materials highly affected compressibility and water permeability [26]. These factors should be considered to examine the mechanical behaviors (e.g., settlement, consolidation) and mass transport processes in the waste body.

Gas Transport Parameters and Pore Tortuosity Characteristics
Measured gas diffusivity (D p /D 0 ) of the tested samples was shown as a function of air-filled porosity (ε) in Figure 6a. For incinerated bottom ash (<2 mm and <9.5 mm) and a landfill cover soil (<2 mm and <35 mm) [13], the D p /D 0 increased with increasing of ε and the data mostly ranged in power-law D p /D 0 (ε) models of Buckingham (1904) [30] and Marshall (1959) [31]. On the other hand, the D p /D 0 values of buried waste (<2 mm and <9.5 mm) gave an increment at ε = 0~0.3, however, the D p /D 0 became almost constant at ε > 0.3 and the data ranged between Buckingham model (ε 2 ) and D p /D 0 = ε 3 . Measured air permeability (k a ) of the tested samples were shown as a function of ε in Figure 6b. In Figure 6b, an empirical power functional k a model with b = 1.5 (Marshal's tortuosity factor [31]) was adopted to capture the range of measured values [13]. Contrary to the relations of D p /D 0 (ε) (Figure 6a), the k a values for tested samples did not show a clear increment with increasing of ε. Especially, the measured k a of buried waste (<2 mm and <9.5 mm) became almost constant irrespective of ε values. The measured k a (ε) for incineration ash (<2 mm and <9.5 mm) and landfill cover soil (<2 mm and <35 mm) mostly ranged between 2000ε 1.5 and 70ε 1.5 .
Environments 2021, 8, x FOR PEER REVIEW 10 of 14 Figure 6. (a) Gas diffusivity, Dp/D0, as a function of air-filled porosity, ε (b) Air permeability, ka, as a function of ε for incinerated bottom ash, buried waste, and landfill cover soil [13].
It was reported that the compaction effort and packed density highly affected Dp/D0(ε) and ka(ε) of geomaterials and soils [10,13,32]. The particle rearrangement and water blockage due to the compaction change the pore structural network in the compacted samples, causing the change of pore structural parameters such as pore size distribution, pore tortuosity and connectivity [33]. In this study, two empirical indices for characterizing the change of pore structural parameters induced by the compaction were used to understand the pore tortuosity and connectivity based on the measured Dp/D0 values. One was the so-called pore connectivity-tortuosity factor, Xg, in the Buckingham model [30]: The other one was called the pore tortuosity factor, T, in the tortuosity-based Dp/D0 model k a = 70 ε 1.5 Figure 6. (a) Gas diffusivity, D p /D 0 , as a function of air-filled porosity, ε (b) Air permeability, k a , as a function of ε for incinerated bottom ash, buried waste, and landfill cover soil [13].
It was reported that the compaction effort and packed density highly affected D p /D 0 (ε) and k a (ε) of geomaterials and soils [10,13,32]. The particle rearrangement and water blockage due to the compaction change the pore structural network in the compacted samples, causing the change of pore structural parameters such as pore size distribution, pore tortuosity and connectivity [33]. In this study, two empirical indices for characterizing the change of pore structural parameters induced by the compaction were used to understand the pore tortuosity and connectivity based on the measured D p /D 0 values. One was the so-called pore connectivity-tortuosity factor, X g , in the Buckingham model [30]: (1) The other one was called the pore tortuosity factor, T, in the tortuosity-based D p /D 0 model proposed by Moldrup et al. (2001) [34]: The X g and T values were plotted as a function of ε, as shown in Figure 7. For tested samples including landfill soil, the X g values ranged mostly between 1 to 3. The X g values incinerated bottom ash (<2 mm and <9.5 mm) and landfill cover soil (<2 mm and <35 mm) did not show any significant increment in X g (ε). For the buried waste (<2 mm and <9.5 mm); however, the X g values increased with an increase of ε at ε > 0.3 and reached X g ∼ = 3. This increment in X g corresponds to no significant increase in D p /D 0 (ε) at ε > 0.3 in Figure 6a, indicating that the compaction of buried waste at the dry condition (high ε) caused more tortuous-connected pores compared to the compacted samples at the wet condition (low ε). On the other hand, the T values decreased with the increase of ε for all tested samples up to around ε = 0.3, including a landfill cover soil shown in Figure 7b, and ranged between 1.2 to 4. At ε > 0.3, the T values became constant irrespective of ε. This strongly suggests that the pore tortuosity of samples depends on the moisture condition at the compaction and the pore network became less tortuous (smaller T) for the samples compacted at the drier condition. did not show any significant increment in Xg(ε). For the buried waste (<2 mm and <9.5 mm); however, the Xg values increased with an increase of ε at ε > 0.3 and reached Xg ≅ 3. This increment in Xg corresponds to no significant increase in Dp/D0(ε) at ε > 0.3 in Figure  6a, indicating that the compaction of buried waste at the dry condition (high ε) caused more tortuous-connected pores compared to the compacted samples at the wet condition (low ε). On the other hand, the T values decreased with the increase of ε for all tested samples up to around ε = 0.3, including a landfill cover soil shown in Figure 7b, and ranged between 1.2 to 4. At ε > 0.3, the T values became constant irrespective of ε. This strongly suggests that the pore tortuosity of samples depends on the moisture condition at the compaction and the pore network became less tortuous (smaller T) for the samples compacted at the drier condition. Finally, the relations of T and ε were fitted with a power function, as shown in Figure  8. Except for some irregular plots in Figure 8a,c, the power function well fitted the T(ε) relations with high r 2 values. The fitted T(ε) relations can be directly combined with Equa- Finally, the relations of T and ε were fitted with a power function, as shown in Figure 8. Except for some irregular plots in Figure 8a,c, the power function well fitted the T(ε) relations with high r 2 values. The fitted T(ε) relations can be directly combined with Equation (2) and be applicable to estimate D p /D 0 (ε) for compacted tested samples [e.g., D p /D 0 = ε(0.72e 0.59 ) −2 ] for incinerated bottom ash). This approach could evaluate the gas diffusion process of compacted dumped materials in landfill sites' waste bodies.

Conclusions
This study investigated the compaction property, compressibility, hydraulic property, and gas transport parameters for compacted waste samples at different moisture

Conclusions
This study investigated the compaction property, compressibility, hydraulic property, and gas transport parameters for compacted waste samples at different moisture conditions by a series of laboratory experiments. It can be concluded that the measured indices for characterizing compaction property, compressibility, and hydraulic property depended highly on the types of waste (waste composition) and varied widely depending on the waste composition. The measured gas diffusivity and air permeability for incinerated bottom ash and landfill cover soil became higher than those of buried waste. Based on the analysis of the pore structural indices, it can be shown that the pore tortuosity of samples depended on the moisture condition at the compaction and the pore network became less tortuous for the samples compacted at the drier condition. Besides, unique power functional relations were found between pore tortuosity and air-filled porosity for the compacted samples. Further studies are needed to examine the scale-up of testing samples (i.e., representative volume size). Especially, the rearrangement of coarser fraction in unburnable waste under compaction process would be dependent on the mold size and the formation of pore networks that affects the gas transport parameters varies depending on the sample size. Additionally, it is essential to apply the measured transport parameters for investigating the gas exchange process and gas transport characteristics in the waste body, fully considering the microbial activities, heterogeneity/variety of dumped waste materials, and repeated wet and dry cycles.