Effect of Particle Size on the Hydraulic Characteristics of Mechanically and Biologically Treated Waste

Mechanical biological treatment (MBT) is a waste processing technology that helps conserve resources and reduce emissions harmful to the environment. 'e treatment of municipal solid waste (MSW) using MBT is a hot topic in environmental geotechnical engineering. Permeability tests were carried out on MBT waste using a compression and permeability combined apparatus and a large-scale vertical permeability apparatus taking the influence of particle size into consideration. 'e permeability of samples with smaller particle sizes was found to be lower for the same pressure and dry mass (%) of component. 'e best-fit line between the logarithmic permeability and variables such as the dry density was linear. As the dry density increased or the void ratio decreased, the permeability of samples with smaller particles decreased more. 'e logarithmic permeability increased with the increase in the average particle size and void ratio.'e permeabilities of MBTwaste corresponding to particle size ranges of 0–10, 0–20, and 0–40mm were 10–10, 10–10, and 10–10m/s, respectively. 'e difference between MBT waste and MSW was analyzed in terms of their permeability. 'e results of MBT waste were compared with those reported in previous studies to provide reference for the permeability analysis of MBT landfills.


Introduction
Mechanical biological treatment (MBT) has attracted increased research interest since it was proposed in the late 1990s [1,2]. e MBT technology involves screening and shredding by mechanical treatment, selecting large-sized plastic by hand and recyclable materials by magnetic separation, and finally treating the organic materials by biological treatment; the remaining waste is disposed to landfills [3,4]. Converting waste to a biostable and low-moisture MBT waste can help decrease the volume, leachates, and gas content in landfills, thus reducing the harmful effects of methane on the environment [5,6].
MBT waste has varied characteristics because it contains municipal solid waste (MSW) from different regions. Adani et al. [7] found that in Italy the input raw materials (MSW) are treated differently in the MBT process. e waste (<50 mm) is biologically treated and later landfilled after decreasing the particle size to less than 20 mm through mechanical sieving and biological recycling. Zhang et al. [8] obtained different grain size distribution of MBT waste (from UK) with different shredding treatment. Kuehle-Weidemeier [9] found the uniformity coefficients of MBT waste (from Germany) with particle size ranges of 0-60, 0-40, and 0-20 mm to be 50, 123, and 192, respectively. In another study, Tungtakanpoung [10] determined different contents of the components of MBT waste (from ailand) with three particle size ranges (<10, 10-40, and >40 mm); the results were quite different from the results of MBT waste in Italy [11]. e particle size significantly influences the permeability of MBT waste. Ziehmann et al. [12] found that the permeabilities of MBT waste with particle size ranges of 0-60, 0-40, and 0-30 mm are in the ranges of 4.35 × 10 −8 -1.63 × 10 −5 , 2.6 × 10 −8 -1.04 × 10 −5 , and 3.76 × 10 −8 -6.61 × 10 −6 m/s, respectively. Siddiqui [13] found the ranges of the permeability of MBT waste with particle size ranges of 0-60 and 0-20 mm to be 5.3 × 10 −7 -6.6 × 10 −5 and 8 × 10 −7 -7.8 × 10 −5 m/s, respectively. Table 1 lists the permeabilities of MBT waste with different particle sizes. e relationship between the permeability and the pressure of MBT waste (Table 1) with one particle size has been studied by several researchers [9,18,19]. In addition, the relationship between the permeability and the dry density of MBT waste with one particle size has also been studied [20,21]. In this study, a systematic analysis on the MBT waste (from Hangzhou, China) with different particle size ranges (0-10, 0-20, and 0-40 mm) was carried out using a compression and permeability combined apparatus and a large-scale vertical permeability apparatus. e relationships between the permeability and the variables such as the dry density were established taking the impact of particle size into consideration. e hydraulic characteristics were analyzed by comparing the wastes in relevant countries. Figure 1 shows the MBT process of MSW adopted from a pilot MBT project at the Tianziling Landfill (in Hangzhou, China) in the subtropical monsoon climate zone, which is warm and humid [23]. Plastic, textiles, and metal were recovered by mechanical treatment, the remaining waste was hydrolyzed, and the MBT waste was kept in the workshop for drying treatment. Note. σ (kPa) pressure; ρ d (g/cm 3 ) dry density; D (mm) specimen diameter; k (m/s) permeability.

Materials.
e materials with the quantities of almost 70 barrels were obtained from several workshops and transported to the environmental geotechnical laboratory. e capacity of each barrel was 120 L. Samples of the materials used for the tests were selected using the quartile method. ree round sieves with sizes of 10, 20, and 40 mm were, respectively, placed on three top dual-purpose screen shakers and the original MBT waste was placed in the sieves. A portion of the materials was passed through the sieves after shaking for 15-20 min and the samples of P1-1 (0-10 mm), P1-2 (0-20 mm), and P1-3 (0-40 mm) were, respectively, obtained from the materials under the sieves. After preparing samples P1-1, P1-2, and P1-3, three round sieves with sizes of 10, 20, and 40 mm were also placed on three top dualpurpose screen shakers, respectively. e original waste was also placed in the sieves, then crushed, and shaken until all the materials in the sieve passed through their respective round sieves. Samples of P2-1 (0-10 mm), P2-2 (0-20 mm), and P2-3 (0-40 mm) were, respectively, obtained from the materials under the sieves. e type of components in the samples (P1-1, P1-2, and P1-3) is the same as that in the original MBT waste; however, the dry mass (%) of the components is different. e types and dry mass (%) of components in the samples P2-1, P2-2, and P2-3 and the original MBT waste are the same. erefore, the dry mass (%) of components in the samples P1-1, P1-2, and P1-3 and that in the samples P2-1, P2-2, and P2-3 are different. Figure 2 shows the type and dry mass (%) of components in samples P1-1 (0-10 mm), P1-2 (0-20 mm), and P1-3 (0-40 mm) and the original MBT waste (same as samples P2-1, P2-2, and P2-3) obtained from the component analysis. e unidentified component is the one that could not be identified or crushed, whereas fine components are those that are finer than 5 mm. e P1-1 sample contained fines (48.4%) and wood (25.8%). e P1-2 sample contained fines (28.5%), rubber and plastic (20.8%), wood (18.7%), textiles (12.4%), and glass (12.3%). e P1-3 sample contained rubber and plastic (24%), wood (14.8%), textile (14.7%), and glass (14.7%). e original MBT waste contained rubber and plastic (30.9%), textile (16.8%), fines (16%), wood (11.5%), and glass (11.3%). P1-1 had the highest fine component content, whereas the original MBT waste had the least. However, the rubber and plastic component content in P1-1 was the least, whereas the original MBT waste had the highest. e metal components in the P1-1, P1-2, P1-3, and the original MBT waste were less. Plastics and textiles accounted for a larger proportion of the MBT waste samples with larger particle sizes; fines accounted for a larger proportion with smaller particle sizes. us, plastics and textiles were mainly of a large size, and the fines were mainly fine particles. A correlation exists between the dry mass (%) of the component and the particle size.

Testing Apparatus.
Compression and permeability related tests were conducted to determine the permeability of MBT waste with smaller particle sizes (0-10 and 0-20 mm) using a compression and permeability combined apparatus considering the effects of pressure (Figure 3(a)) [24]. A maximum vertical pressure of 1600 kPa could be applied to the permeation column using a lever. e diameter and height of the permeation column were 150 and 347 mm, respectively. A cylindrical header tank was fixed on an adjustable plate with a maximum height of 1200 mm. e hydraulic gradient was set by varying the height of the header tank. Large-scale vertical permeability tests were conducted to determine the permeability of MBT waste with larger particle sizes (0-40 mm) using a large-scale vertical permeability apparatus (Figure 3(b)) [25]. e diameter and height of the permeation column were 400 and 800 mm, respectively. e hydraulic gradient was set by varying the height of the header tank.

Experimental Procedure.
e grain size analysis tests and permeability tests were conducted in accordance with the Chinese Technical Specification for Soil Test of Landfilled Municipal Solid Waste (CJJ/T204-2013) [26]. Permeability tests were conducted on the samples P1-1, P2-1, P1-2, and P2-2 using a compression and permeability combined apparatus taking the effect of pressure into consideration. Permeability tests were conducted on the samples P1-3 and P2-3 using a large-scale vertical permeability apparatus to avoid the scale effect. For this study, 540 groups of permeability tests were conducted and the materials with the quantities of almost 5 barrels were processed.
Consider sample of P1-1 under a pressure of 50 kPa with hydraulic gradients of 2, 1.75, 1.5, 1.25, 1, and 0.5 as an example to explain the method of the test. e sample was selected using the quartile method and placed in a constant temperature oven at 65°C for drying. e dried sample was placed into the permeability column five times and compacted to ensure a uniform sample. Filter paper and porous stone were placed at both ends of the sample. e total mass (M 1 ) and initial height (H 1 ) of the sample were recorded. e permeability column was fixed after pushing it under the cover. Water flowed from the head tank to the inlet. After achieving a stable outflow, the sample was saturated. en, a pressure of 50 kPa was set and the hand wheel was adjusted to keep the lever level. e permeability test was conducted after compressing the sample for 24 h. e settlement (S) was recorded, and the height of the sample compressed for 24 h was calculated using equation e hydraulic gradient was set as 2, and the height of the header tank was adjusted. ree parallel tests were performed after the flow of water became stable. e hydraulic gradients were reduced to 1.75, 1.5, 1.25, 1, and 0.5. After the test with six hydraulic gradients was completed, the weight was removed, and the sample was taken out.
e above steps were repeated to refill the sample of P1-1, and the following tests were conducted under pressures of 100, 200, 300, and 400 kPa, respectively. e data were recorded until all the tests have been completed.
Consider sample P1-3 as an example to explain the method of the test. e sample was selected using the quartile method and placed in a constant temperature oven at 65°C for drying. e dried sample was placed into the permeability column five times and compacted to ensure a uniform sample. Filter paper and porous stone were placed at both ends of the sample. e total mass (M 2 ) and initial height (H 2 ) of the sample were recorded. e water flowed from the head tank to the inlet. After achieving a stable outflow, the sample was saturated. An initial hydraulic gradient of 1.25 was set, and the height of the header tank was adjusted. ree parallel tests were conducted after the flow of water became stable. e hydraulic gradients were reduced to 1, 0.75, 0.5, 0.25, and 0.1. After the test with six hydraulic gradients has been completed, the sample was taken out. e above steps were repeated to refill the sample of P1-3, and the dry density was varied by increasing the Advances in Civil Engineering compaction and stacking weight. Tests were conducted on P1-3 with dry densities of 0.150, 0.154, 0.166, 0.205, and 0.223 g/cm 3 , whereas tests were conducted on P2-3 with dry densities of 0.299, 0.339, 0.357, 0.400, and 0.417 g/cm 3 . 180 groups of permeability tests were conducted and the samples with the quantities of almost 3 barrels were processed. e data were recorded until all the tests were completed. Figure 4 shows the grain size distribution curves under different particle sizes. e percentages of P1-1, P2-1, P1-2, P2-2, P1-3, P2-3, and the original MBT waste that passed through a 5 mm sieve were 36.0, 26.9, 26.8, 25.2, 24.7, 19.5, and 16.0, which decreased in this order. Table 2 lists the characteristic particle sizes (d 10 , d 30 , d 50 , and d 60 , which are diameters of the particle size at which 10%, 30%, 50%, and 60% of the materials pass through the sieve), uniformity coefficient (C u ), and curvature coefficient (C c ) of P1-1, P1-2, P1-3, P2-1, P2-2, P2-3, and the original MBT waste. e uniformity coefficients of P1-1, P1-2, P1-3, P2-2, P2-3, and the original MBT waste are greater than 5, and the curvature coefficient is in the range of 1-3. Samples of P1-1, P1-2, P1-3, P2-2, P2-3, and the original MBT waste are well-graded waste, whereas P2-1 is not. Figure 5 shows the relationship between the permeability and pressure of P1-1, P2-1, P1-2, and P2-2. e best-fit line between the logarithmic permeability (lgk) and the pressure (σ) is linear (equation (1)). Table 3 lists the coefficients A and B and the correlation coefficient R 2 .

Relationship between Permeability and Pressure.
lg k � A + Bσ. (1) As shown in Figure 5, the permeabilities of P2-1 and P2-2 are in the range of 10 −4 -10 −3 m/s, whereas the permeabilities of P1-1 and P1-2 are in the range of 10 −5 -10 −4 m/s under a pressure of 0 kPa. Furthermore, the permeabilities of P2-1 and P2-2 are in the range of 10 −10 -10 −7 m/s, whereas the permeabilities of P1-1 and P1-2 are in the range of 10 −8 -10 −6 m/s under a pressure of 400 kPa. erefore, the permeabilities of P2-1 and P2-2 decreased to a greater extent with increasing pressure, and the slope of the best-fit line of P2-1 and P2-2 is greater than that of P1-1 and P1-2. e permeability of P2-1 became lower than that of P1-1 when the pressure increased slightly above 150 kPa. e permeability of P2-2 became lower than that of P1-2 at pressures that are slightly below 150 kPa. As a result of the heterogeneity and high compressibility of the MBT waste, the main components had an effect on permeability. Scanning electron microscope (SEM) images (see Figure 6) of the components of plastic and textile in the MBT waste were obtained at a magnification of 500. As the plastic surface has no pores, the large-sized plastic changed the vertical seepage path and reduced the vertical permeability under compaction conditions [16]. e textile was largely filled with impurities, and the pores inside the fiber were significantly reduced. e main reason could be that the samples P2-1 and P2-2 contained more rubber, plastics, and textiles, but less fines. e compression of rubber plastics and textiles is less at pressures below 150 kPa, and the interparticle voids between the components are relatively large. e degree of compression of samples P2-1 and P2-2 is increased with the increase in the pressure. e components were further compressed under a pressure of more than 150 kPa, and the interparticle voids were relatively reduced.
As the best-fit line of P2-1 (0-10 mm) was located below that of P2-2 (0-20 mm), the permeability of P2-1 was lower than that of P2-2 under the same pressure. When the component and content of the sample were the same, the permeability of the smaller particle sizes was lower, which is consistent with results in previous studies [9,14,15]. As the best-fit line of P1-1 (0-10 mm) was located below that of P1-2 (0-20 mm), the permeability of P1-1 was lower than that of P1-2 under the same pressure. e permeabilities of P2-1 and P2-2 were slightly lower than those of P1-1 and P1-2 under a pressure of 400 kPa. ere were differences in the particle size and component content in the samples, indicating that the influence of particle size on the permeability is less than that of the component content under a pressure of 400 kPa. Figure 7 shows the relationship between the permeability and the dry density of samples with different particle sizes. e best-fit line between the logarithmic permeability (lgk) and the dry density (ρ d ) is linear (equation (2)). Table 4 lists the coefficients C and D and the correlation coefficient R 2 .

Relationship between Permeability and Dry Density.
lg k � C + Dρ d .
(2) Figure 7 shows that the absolute values of the slopes of the best-fit lines (P1-1, P1-2, and P1-3) decrease sequentially with the increase in the particle size of the samples; similar results were obtained for P2-1, P2-2, and P2-3. e permeability of the samples with smaller particle sizes decreased to a greater extent with the increase in the dry density.
e absolute values of the slopes of the best-fit lines between P1-1 and P2-1 were essentially similar. e absolute value of the slope of the best-fit line of P1-2 was less than that of P2-2. e absolute value of the slope of the best-fit line of P1-3 was much lower than that of P2-3. e differences between the slopes of the samples (P1-1 and P2-1, P1-2 and P2-2, and P1-3 and P2-3) increased at increasing particle size. is may be due to the different contents in the samples, given that the particle size and component were the same. e difference in the interparticle voids increased with an increase in the particle size. e permeability ranged from 4.9 × 10 −3 to 5.24 × 10 −10 m/s at a dry density range of 0.15-0.84 g/cm 3 . As the dry density of samples P1-1, P1-2, and P1-3 was 0.7 g/cm 3 , the permeability intervals of P1-2 (10 −7 -10 −6 m/s) and P1-3 (10 −5 -10 −4 m/s) were lower than that of P1-1 (10 −4 -10 −3 m/s). Xie et al. [17] obtained a similar result; namely, the permeability of the MBT waste with a particle 6 Advances in Civil Engineering  Advances in Civil Engineering size range of 0-20 mm was lower than that of the MBT waste with a particle size range of 0-6.3 mm for the same dry density. Figure 8 shows the relationship between the permeability and the void ratio of samples with different particle sizes. e best-fit line between the logarithmic permeability (lgk) and the void ratio (lge) is linear (equation (3)). Table 5 lists the coefficients E and F and the correlation coefficient R 2 .

Relationship between Permeability and Void Ratio.
lg k � E + F lge.
(3)   permeability of the samples with smaller particle sizes decreased to a greater extent at a decreasing void ratio. e best-fit line of P1-1 was above that of P2-1 in the void ratio range of 0.6-0.9, indicating that the permeability of P1-1 was greater than that of P2-1. e best-fit line of P1-2 was above that of P2-2 in the void ratio range of 0.7-1.6, indicating that the permeability of P1-2 was greater than that of P2-2. e interparticle voids were similar with the same void ratio. e main reason could be that the content of plastic in samples P2-1 and P2-2 was more than that in samples P1-1 and P1-2 for the same void ratio. e vertical seepage path was prolonged because of the large-sized plastics, and the corresponding permeability decreased [17].

Relationship between Permeability and Average
Particle Size. Biogas produced from the degradation of organic materials reduces the permeability of MSW [27]. However, the content of organic materials in the MBT waste was relatively low, with less biogas to be produced. e permeability decreased mainly with the decrease in the void ratio. e distributions of the void ratio and permeability were different at different average particle sizes. Figure 9 shows the relationships between the permeability, void ratio, and average particle size. e figure shows that the permeability increases with the increase in the average particle size and void ratio in the samples P1-1, P1-2, and P1-3 as well as samples P2-1, P2-2, and P2-3. e permeability of P1-1 ranged from 10 −8 to 10 −5 m/s with an average particle size of 5.4 mm and a void ratio range of 0.65-0.84. e permeability of P1-2 ranged from 10 −7 to 10 −4 m/s with an average particle size of 9.6 mm and a void ratio range of 0.8-1.7. e permeability of P1-3 ranged from 10 −4 to 10 −3 m/s with an average particle size of 13.8 mm and a void ratio range of 5.1-8.2. is may be due to the increase in the average particle size, decrease in the content of fines in samples P1-1, P1-2, and P1-3, and decrease in the interparticle voids in samples with smaller particle sizes. e multiple independent seepage channels gradually expanded into a single seepage channel, and the seepage channel widened significantly. e seepage path was significantly reduced, and the permeability increased.   e permeability of P2-1 ranged from 10 −10 to 10 −5 m/s with an average particle size of 5.9 mm and a void ratio range of 0.66-0.82. e permeability of P2-2 ranged from 10 −8 to 10 −4 m/s with an average particle size of 9.4 mm and a void ratio range of 0.7-1.8. e permeability of P2-3 ranged from 10 −5 to 10 −3 m/s with an average particle size of 12.4 mm and a void ratio range of 2.3-3.6. e type and dry mass (%) of the components in the samples P2-1, P2-2, and P2-3 were the same, and the increase in the average particle size produced an equivalent effect as the size. e uniformity coefficients of samples P2-1, P2-2, and P2-3 increased with the increase in the average particle size. e larger particle size samples had greater heterogeneity among the components such as plastic and textile. e permeability increased more with larger particle sizes.

Comparison of the Results of MBT Waste Samples in China and Some Countries.
e difference in the particle size of the waste was found to be significant. e relationship between permeability and the dry density of MSW with different particle sizes was obtained. e permeability of samples with smaller particle sizes was found to be lower. e main reason was that the particle size decreases with degradation [28]. Compared with MSW, the degradable components in the MBT waste were less, and the particle size was mainly related to the MBT process. Figure 10 shows a comparison of the permeability results of the MBT waste. e permeability gradually increases with an increase in the dry density, and the best-fit line between the logarithmic permeability and the dry density is linear. e absolute values of the slope of best-fit line of the samples (Kuehle-Weidemeier [9]; Siddiqui [13]; P1-2) are between those of the samples P1-1 and P1-3.
is indicates that the decrease in the permeability of the samples with size ranges of 0-20 and 0-30 mm was greater than that of the samples with a size range of 0-40 mm and less than that of the samples with a size range of 0-10 mm as the dry density increased. e best-fit line (Siddiqui [13]) was located below the best-fit line of P1-2 in the dry density range of 0.45-0.65 g/cm 3 , indicating that the permeability (Siddiqui [13]) was low. e main reason could be the different grain size distributions. e average particle size (Siddiqui et al. [29]) was smaller than that of P1-2. e components with a particle size less than 5 mm in the sample (Siddiqui et al. [29]) were more than those in the sample P1-2. e best-fit line (Kuehle-Weidemeier [9]) almost coincided with the best-fit line of P1-2, indicating that the hydraulic characteristics were similar. e main reason was that the average particle size (Kuehle-Weidemeier [9]) was lower than that of P1-2, and the uniformity coefficients were similar. In summary, the degree of uniformity and the distribution of the pores were similar. Figure 11 shows a comparison of the permeability of MBT waste and MSW considering the effects of dry density and average particle size. e distribution range of the MBT waste (Kuehle-Weidemeier [9]; Xie et al. [17]; Siddiqui [13]; this paper) is below that of MSW (Reddy et al. [28]; Jie et al. [30]), indicating that the permeability of MBT waste is less than that of MSW. e particle size of the MBT waste was reduced by mechanical treatment, and the content of the organic materials was reduced through biological treatment. e permeability of the MBT waste and MSW decreased with the increase in the dry density for the same average particle size; the permeability of MBT waste with a lower dry density and that of MSW with a higher dry density were the same. e distribution range of the permeability of MBT 10 Advances in Civil Engineering waste and MSW generally increased with the increase in the average particle size at the same dry density.

Conclusions
Permeability tests were conducted on MBT waste (obtained from the Tianziling Landfill in Hangzhou) with different particle sizes using a compression and permeability combined apparatus and a large-scale vertical permeability apparatus. e following conclusions can be drawn from the study results: (1) e components in the waste samples were related to the particle sizes. e components of plastics and textiles in the MBT waste were mainly of large size. Interface MBT waste, this paper MBT waste, Siddiqui [13] MBT waste, Kuehle-Weidemeier [9] MBT waste, X ie et al. [17] Landfilled waste, Jie et al. [30] Landfilled waste, Reddy et al. [28] Fresh waste, Reddy et al. [28]  Samples of P1-1, P1-2, P1-3, P2-2, P2-3, and the original MBT waste were well-graded waste, whereas P2-1 was not. (2) e permeabilities of the samples of P1-1, P2-1, P1-2, and P2-2 decreased as the pressure increased, and a pressure of 150 kPa was the critical value. For the same component content and pressure, the permeability of samples with a smaller particle size was lower. (3) As the particle size increased, the difference in the slopes of the best-fit lines between the logarithmic permeability and the dry density of the samples (P1-1 and P2-1, P1-2 and P2-2, and P1-3 and P2-3) increased. At the same dry density (0.7 g/cm 3 ), the permeability of P1-1 was 1-2 orders of magnitude greater than those of P1-2 and P1-3. e permeability of samples with smaller particle sizes decreased to a greater extent with the increase in the dry density. (4) With the increase in the particle size, the absolute values of the slopes of the best-fit lines between the logarithmic permeability and the logarithmic void ratio of samples P1-1, P1-2, and P1-3 decreased sequentially and those of samples P2-1, P2-2, and P2-3 decreased in this order as well. As the void ratio decreased, the permeability of samples with a smaller particle size decreased to a greater extent. (5) As the average particle size increased, the permeability and void ratio increased. e permeability of samples P1-1, P1-2, and P1-3 ranged from 10 −8 to 10 −5 , 10 −7 to 10 −4 , and 10 −4 to 10 −3 m/s, with average particle sizes of 5.4, 9.6, and 13.8 mm and void ratio ranges of 0.65-0.84, 0.8-1.7, and 5.1-8.2, respectively. (6) A comparison of the results of MBT waste from China with those from other countries shows that the absolute values of the slopes of the best-fit lines (0-20 and 0-30 mm) were between those of the best-fit lines (0-10 and 0-40 mm) at increasing dry density. A comparison of the permeability of MBT waste with that of MSW shows that the distribution of the permeability of MBT waste was below that of MSW, and the permeability was relatively low.

Data Availability
e data used to support the findings of this study are included within the article.

Conflicts of Interest
e authors declare that they have no conflicts of interest regarding the publication of this paper.