56565
research-article2015
Original Article
Parameter determination of a compression model for landfilled municipal solid waste: An experimental study
Waste Management & Research 2015, Vol. 33(2) 199–210 © The Author(s) 2015 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/0734242X14565657 wmr.sagepub.com
Xiao Bing Xu , Tony Liang Tong Zhan, Yun Min Chen and Qi Gang Guo
Abstract The methods to determine the parameters of a one-dimensional compression model for landfilled municipal solid waste were investigated. In order to test the methods for parameter determination, long-term laboratory compression experiments were carried out under different surcharge loads (i.e. 100, 200 and 400 kPa). Based on the measured compression strain and the reported creep index, the modified primary compression indexes, pre-consolidation pressure and ultimate biodegradation-induced secondary compression strain were determined using the proposed methods. It was found that the simulated compression could not capture the measured secondary compression behavior when using a constant value of biodegradation-induced compression rate coefficient. The variation of the rate coefficient with the change of decomposition rate should be considered during modeling. Accordingly, the biodegradationinduced secondary compression strain in the compression model should be expressed in an incremental form in order to consider the variation of the rate coefficient. Keywords Biodegradation, compression, creep, landfill, municipal solid waste, settlement, stress
Introduction When municipal solid waste (MSW) is placed in a landfill, settlement continues over an extended period of time, with a final settlement reaching 25–50% of the initial fill height (Stearns, 1987; Ling et al., 1998). The accurate prediction of settlement would allow estimation of landfill capacity, planning of construction sequence, design of both intermediate and final covers, design of piping systems and planning for expansions (Hunte et al., 2007). Mechanisms of settlement in landfills are complex due to the extreme heterogeneity, compressibility and biodegradability of MSW. For an unsaturated waste, predominant mechanisms resulting in settlement differ in distinguishable stages which generally consist of: (i) primary compression resulting from the compression or crushing of compressible matters responding to the surcharge of subsequently placed MSW, and (ii) long-term secondary compression due to mechanical creep and biodegradation. During secondary compression, creep is caused by the time-dependent particle distortion (i.e. bending, crushing), particle reorientation and raveling (Dixon and Jones, 2005). Biodegradation-induced compression is a consequence of the collapse due to a change in strength and the degradation of organic compounds. Primary compression is commonly determined through oedometer tests and described by the modified primary compression index C’c (i.e. the slope of compression strain and logarithmic stress curve). Dissipation of water and gas within the
pore space of MSW is rapid upon loading due to the unsaturated nature of MSW (El-Fadel and Khoury, 2000). Therefore, the time for the completion of primary compression under a specific load is generally within a few days (Bareither et al., 2012). The value of C’c for MSW falls in a wide range from 0.08 to 0.46 in reported studies (Chen et al., 2009; Reddy et al., 2009; Bareither et al., 2012). Experimental studies indicate that C’c of MSW is mainly affected by waste composition, particle size and degree of degradation (Hossain and Gabr, 2009). A decrease of C’c following sustained loading and waste decomposition, relative to fresh MSW, was reported by Chen et al. (2010) and Bareither et al. (2012). Through long-term laboratory compression tests with biodegradation inhibited, the creep index Cα (i.e. the slope of compression strain and logarithmic time curve) of MSW was determined (Hossain et al., 2003; Ivanova et al., 2008; Chen et al., 2010; Siddiqui et al., 2013). Laboratory study carried out by Hossain and
MOE Key Laboratory of Soft Soils and Geoenvironmental Engineering, Institute of Geotechnical Engineering, Zhejiang University, Hangzhou, China Corresponding author: TLT Zhan, MOE Key Laboratory of Soft Soils and Geoenvironmental Engineering, Institute of Geotechnical Engineering, Zhejiang University, Yuhangtang Road 866#, Hangzhou 310058, China. Email: [email protected]
Downloaded from wmr.sagepub.com at The University of Iowa Libraries on June 5, 2015
200
Waste Management & Research 33(2) σ σ2
σ1 O A1 A2
t0
t1
Ai
ε(σiˈtj)
t2
tj-1
tj
t
B1 B2
Ai-1
well as a field monitoring project, the effect of decomposition rate change on biodegradation-induced compression was highlighted.
σi
σi-1
σ1
C2 C3
Xi-2 Xi-1
σ2 Yi-1 Yi
Zi
σi-1 σi
Model description Based on the compression model of Chen et al. (2010), a modified compression model expressing the total compression strain (ε) of MSW as a sum of primary compression, creep and biodegradationinduced compression strains was proposed:
ε
ε = ε p + ε sc + ε sb (1)
σ ε p = CC′ 0 log10 (2) σ0
Figure 1. Working principle of the 1-D compression model (Chen et al., 2010).
Gabr (2009) indicated that Cα of MSW with different degrees of biodegradation and values of maximum particle size were similar. Bareither et al. (2013) found that Cα of fresh and decomposed MSW was not affected by stress and specimen size. A relatively narrow range of about 0.01–0.05 for Cα was reported for MSW from these published studies. Many long-term laboratory and field studies on MSW settlement were carried out to understand biodegradationinduced compression. More satisfactory simulation for measured biodegradation-induced compression has been found on the basis of first-order biodegradation kinetics (Ivanova et al., 2008; Elagroudy et al., 2008; Gourc et al., 2010). The potential of biodegradation-induced compression strain was reported in a range of 0.13–0.27 for fresh MSW (Chen et al., 2010), and was believed to increase with increasing content of degradable matter. However, the determination of the rate of biodegradation-induced compression is still a challenge for researchers and engineers. Ivanova et al. (2008), Bareither et al. (2013) and Siddiqui et al. (2013) conducted pioneer studies on the relationship between the degrees of biodegradation-induced compression and biodegradation for MSW. Chen et al. (2010) proposed a one-dimensional (1-D) compression model which took two important compression characteristics of MSW into account: (i) the decrease of C’c due to biodegradation and (ii) the load dependence of secondary compression. This compression model was later applied by Chen et al. (2012) and Li et al. (2013) to analyze settlement and storage capacity of MSW landfills. However, in the model of Chen et al. (2010), creep and biodegradation-induced secondary compression were combined and expressed by first-order kinetics. In order to take a closer investigation of the relationship between biodegradation-induced compression and biodegradation process, a modified model separating out the effect of creep on secondary compression was proposed based on the model of Chen et al. (2010). Furthermore, the methods to determine the model parameters were proposed in this paper to make the modified model more applicable for engineers. Long-term laboratory compression experiments were carried out to test the proposed methods. Through simulation of compression data measured from the laboratory experiments as
t ε sc = Cα log10 tp
(3)
ε sb = ε sb∞ 1 − exp − c (t −tb ) (4)
where: εp represents the primary compression strain of waste responding to the surcharge load of σ (kPa); εsc and εsb are the creep and biodegradation-induced compression strains at the time t since the application of σ, respectively; σ0 (kPa) represents the pre-consolidation pressure; C’C0 represents the modified primary compression index for fresh waste; Cα is the creep index for fresh waste; tp (day) is the time for the completion of primary compression; εsb∞ is the ultimate biodegradation-induced secondary compression strain of the fresh waste under the applied load of σ; tb (day) is the time for the start of biodegradation; and c (day-1) is the rate coefficient of biodegradation-induced compression. The working principle of the modified model can be explained by a series of iso-stress compression curves (Figure 1) which represent the time-dependent compression under a constant stress. The real “loading” path of a specific waste layer within a landfill generally follows the path with respect to the applied load and elapsed time, e.g., OA1B1B2C2C3···Xi-2Xi-1Yi-1YiZi. Four assumptions are made including: (i) biodegradation rate of MSW is independent of the stress level; (ii) creep is independent of the stress level and the degree of biodegradation; (iii) the time for the completion of primary compression under a specific stress is short; and (iv) primary compressibility depends on the degree of biodegradation. Based on these assumptions, the total strain ε(σi,tj) at the point Zi is insensitive to the alternative “loading” paths, but depends only on the final applied load (i.e. σi) and the degree of biodegradation (i.e. the elapsed time tj). Therefore, ε(σi,tj) can be calculated by following the two-step “loading” path from point O to point Ai and then to point Zi. As shown in Figure 1, the “loading” path from point O to point Ai (i.e. load from zero to σi) is a monotonic loading applied on the fresh waste deposited. The subsequent path from point Ai to point Zi is
Downloaded from wmr.sagepub.com at The University of Iowa Libraries on June 5, 2015
201
Xu et al. 10 σ0
1
102
103
σ (kPa)
0 εsb∞(σ0) 20
εp(σ)
εsc(tĄ)
40
εsb∞(σ)
60 T0 T∞-sb T∞
ε (%)
Figure 2. Determination of parameters in the compression model.
a monotonic path with respect to time from zero to tj under the constant load of σi.
Determination of model parameters In Equation (2), two parameters including C’C0 and σ0 need to be determined. As shown in Figure 2, the value of C’C0 can be determined by the primary compression curve (i.e. line T0) obtained through laboratory 1-D mechanical compression tests on fresh waste. It is assumed that primary compression strain is zero if the applied load is smaller than σ0. Therefore, the stress at the intersection point of the primary compression curve for fresh waste and the zero strain coordinate is defined as σ0. In Equation (3), two parameters including Cα and tp need to be determined. The value of Cα can be determined from long-term laboratory compression tests on fresh waste under certain surcharge load with biodegradation inhibited. The time with respect to the inflexion point of the compression strain and logarithmic time curve curve for determining Cα is defined as tp, which means long-term compression strain becomes linear with logarithm of time after tp. In Equation (4), three parameters including εsb∞, tb and c need to be determined. A fully decomposed waste can be achieved by performing an accelerated biodegradation test on fresh waste under a sustained load (e.g. σ0). For a given stress σ, as shown in Figure 2, εsb∞(σ) can be determined from the vertical distance between the primary compression curves for fresh waste (i.e. line T0) and fully decomposed waste (i.e. line T∞) minus the creepinduced compression strain (i.e. εsc(t’) = Cαlog10(t’/tp)) caused during the test for achieving the fully decomposed waste. As compared with the time required for achieving the fully decomposed waste, the time required for primary compression test on the fully decomposed waste is assumed to be negligible. Therefore, line T∞-sb determining εsb∞(σ) is parallel to line T∞. εsb∞(σ) can be expressed as follows which indicates that biodegradation-induced compression is dependent on stress level:
σ ε sb∞ ( σ ) = (C ′C∞ − C ′C 0 )log10 + ε sb∞ ( σ0 ) σ0 t′ σ = (C ′C∞ − C ′C 0 )log10 + ε s 0 − Cα log10 σ0 tp
(5)
where: C’C∞ represents the modified primary compression index for the fully decomposed waste; t’ (day) represents the time required for the above accelerated biodegradation test for achieving the fully decomposed waste; εs0 is the total secondary compression strain of the accelerated biodegradation test on the fresh waste under a sustained load of σ0. However, it might be difficult to obtain a fully decomposed waste. The sample with the content of biodegradable matter less than 5% of the initial value of fresh waste is considered to have little variation of C’c after further degradation of the biodegradable matter left. This sample could be used to represent the fully decomposed waste. For a specific landfill, tb can be defined as the time when gas generation starts. The value of c can be obtained from the backanalysis of monitored long-term settlement data at landfills with similar decomposition conditions. Equation (4) can be rewritten as follows based on which the value of c can be determined by plotting ln(εsb∞(σ)- εsb(σ,t)) versus time and fitting the linear curve:
ln ε sb∞ (σ) − ε sb (σ) = −c(t − tb ) + ln ε sb∞ (σ) (6)
Materials and methods Experimental setup A schematic of the 1-D compression apparatus is shown in Figure 3. The specimen cell was made of a stainless steel cylinder with the inner diameter of 200 mm, wall thickness of 10 mm and height of 600 mm. A filter plate was placed at the bottom of the cell and a geotextile was placed above the plate. A steel plate with dead weight placed above was used to produce the compression load on top of the waste specimen. Settlement of the specimen was monitored by a LVDT (5000HCD-200, Weihe Ltd). Two O-rings were set around the load distribution plate and lubricated with vaseline oil to create gas tight seals and minimize the friction between the cell and the load distribution plate. Leachate was collected from the base of the specimen cell. The collected leachate was mixed with sewage sludge and recirculated into the waste specimen through the peristaltic pump in order to enhance biodegradation. The sewage sludge was obtained from the anaerobic digester of a municipal sewage plant in Hangzhou, China. Anaerobic digestion bacteria (e.g. methanogenic bacteria) contained in the sewage sludge could enhance the anaerobic degradation of fresh MSW which generally has little anaerobic digestive bacteria initially (Ivanova, 2007). A porous plate at the bottom of the load distribution plate was used to create uniform distribution of recirculated leachate on the specimen. A gas pressure gauge and a gas collection tube were installed at the top of the load distribution plate. The volume of generated gas was measured by means of the water (i.e. NaHCO3 saturated water) displacement technique. The temperature of the waste specimen was controlled by the means of water bath. A calefaction tube and a cyclic water pump were installed at the bottom of the water tank which
Downloaded from wmr.sagepub.com at The University of Iowa Libraries on June 5, 2015
202
Waste Management & Research 33(2)
Figure 3. Schematic of the compression apparatus. Table 1. Composition of MSW. Composition
Food
Paper
Wood
Textile
Plastic
Metal Glass
Cinder
Others*
(kg kg-1, %, wet basis)
67.7
6.7
0.6
0.6
8.5
1.5
13.7
0.7
*Undistinguishable soil-like matters.
was heat-insulated. The cyclic water pump was used to create a uniform temperature within the water bath. Hartz et al. (1982) found that 41°C was the optimum for the production of methane, and methane production would cease somewhere between 48°C and 55°C. Therefore, a threshold temperature of 41±1°C was set for the calefaction tube to produce optimal temperature condition for waste biodegradation.
obtain a waste specimen with the pre-designed void ratio and unit weight. The predesigned values of void ratio (e = 3.0) and unit weight (γ = 12.17 kN m-3) of the specimen are close to the reported data of waste drilled from the shallow layer of Qizishan landfill, China, respectively (Chen et al., 2009).
Tested materials
Three compression tests were carried out in cell 1, cell 2 and cell 3, respectively. The initial heights of specimens in cells 1, 2 and 3 were 400, 200 and 400 mm, respectively. The amount of waste required for each cell was determined by the pre-designed initial height and the unit weight mentioned above. In order to determine the parameters of the compression model through the proposed methods, each compression test consisted of three stages in time scale: (i) short-term multi-stage compression on fresh waste until reaching the pre-designed sustained load; (ii) continuing long-term compression under the predesigned sustained load with biodegradation enhanced; and (iii) final short-term multi-stage compression on the decomposed waste after secondary compression nearly stopped under the pre-designed sustained load. The measured primary compression strain in stage (i) and (iii) was used to determine C’C0, σ0, C’C∞
Synthetic MSW was made according to the reported composition of fresh MSW generated in Hangzhou, China. As shown in Table 1, degradable matter in waste is mainly composed of food waste (about 67.7%, wet basis) which has a water content of about 85% (He, 2011). Two subsamples of synthetic MSW were used for the determination of physical properties. Waste materials were dried at an oven temperature of 60°C for 48 hours to prevent ignition loss of organic matters, and the water content of the entire specimen (w = 54.6%, wet basis) was measured. The bulk specific gravity (ds = 2.21) of waste was measured using the water replacement method in accordance with the technical code of CJJ176-2012 (MOHURD, 2012). A certain amount of shredded waste (maximum particle size
research-article2015
Original Article
Parameter determination of a compression model for landfilled municipal solid waste: An experimental study
Waste Management & Research 2015, Vol. 33(2) 199–210 © The Author(s) 2015 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/0734242X14565657 wmr.sagepub.com
Xiao Bing Xu , Tony Liang Tong Zhan, Yun Min Chen and Qi Gang Guo
Abstract The methods to determine the parameters of a one-dimensional compression model for landfilled municipal solid waste were investigated. In order to test the methods for parameter determination, long-term laboratory compression experiments were carried out under different surcharge loads (i.e. 100, 200 and 400 kPa). Based on the measured compression strain and the reported creep index, the modified primary compression indexes, pre-consolidation pressure and ultimate biodegradation-induced secondary compression strain were determined using the proposed methods. It was found that the simulated compression could not capture the measured secondary compression behavior when using a constant value of biodegradation-induced compression rate coefficient. The variation of the rate coefficient with the change of decomposition rate should be considered during modeling. Accordingly, the biodegradationinduced secondary compression strain in the compression model should be expressed in an incremental form in order to consider the variation of the rate coefficient. Keywords Biodegradation, compression, creep, landfill, municipal solid waste, settlement, stress
Introduction When municipal solid waste (MSW) is placed in a landfill, settlement continues over an extended period of time, with a final settlement reaching 25–50% of the initial fill height (Stearns, 1987; Ling et al., 1998). The accurate prediction of settlement would allow estimation of landfill capacity, planning of construction sequence, design of both intermediate and final covers, design of piping systems and planning for expansions (Hunte et al., 2007). Mechanisms of settlement in landfills are complex due to the extreme heterogeneity, compressibility and biodegradability of MSW. For an unsaturated waste, predominant mechanisms resulting in settlement differ in distinguishable stages which generally consist of: (i) primary compression resulting from the compression or crushing of compressible matters responding to the surcharge of subsequently placed MSW, and (ii) long-term secondary compression due to mechanical creep and biodegradation. During secondary compression, creep is caused by the time-dependent particle distortion (i.e. bending, crushing), particle reorientation and raveling (Dixon and Jones, 2005). Biodegradation-induced compression is a consequence of the collapse due to a change in strength and the degradation of organic compounds. Primary compression is commonly determined through oedometer tests and described by the modified primary compression index C’c (i.e. the slope of compression strain and logarithmic stress curve). Dissipation of water and gas within the
pore space of MSW is rapid upon loading due to the unsaturated nature of MSW (El-Fadel and Khoury, 2000). Therefore, the time for the completion of primary compression under a specific load is generally within a few days (Bareither et al., 2012). The value of C’c for MSW falls in a wide range from 0.08 to 0.46 in reported studies (Chen et al., 2009; Reddy et al., 2009; Bareither et al., 2012). Experimental studies indicate that C’c of MSW is mainly affected by waste composition, particle size and degree of degradation (Hossain and Gabr, 2009). A decrease of C’c following sustained loading and waste decomposition, relative to fresh MSW, was reported by Chen et al. (2010) and Bareither et al. (2012). Through long-term laboratory compression tests with biodegradation inhibited, the creep index Cα (i.e. the slope of compression strain and logarithmic time curve) of MSW was determined (Hossain et al., 2003; Ivanova et al., 2008; Chen et al., 2010; Siddiqui et al., 2013). Laboratory study carried out by Hossain and
MOE Key Laboratory of Soft Soils and Geoenvironmental Engineering, Institute of Geotechnical Engineering, Zhejiang University, Hangzhou, China Corresponding author: TLT Zhan, MOE Key Laboratory of Soft Soils and Geoenvironmental Engineering, Institute of Geotechnical Engineering, Zhejiang University, Yuhangtang Road 866#, Hangzhou 310058, China. Email: [email protected]
Downloaded from wmr.sagepub.com at The University of Iowa Libraries on June 5, 2015
200
Waste Management & Research 33(2) σ σ2
σ1 O A1 A2
t0
t1
Ai
ε(σiˈtj)
t2
tj-1
tj
t
B1 B2
Ai-1
well as a field monitoring project, the effect of decomposition rate change on biodegradation-induced compression was highlighted.
σi
σi-1
σ1
C2 C3
Xi-2 Xi-1
σ2 Yi-1 Yi
Zi
σi-1 σi
Model description Based on the compression model of Chen et al. (2010), a modified compression model expressing the total compression strain (ε) of MSW as a sum of primary compression, creep and biodegradationinduced compression strains was proposed:
ε
ε = ε p + ε sc + ε sb (1)
σ ε p = CC′ 0 log10 (2) σ0
Figure 1. Working principle of the 1-D compression model (Chen et al., 2010).
Gabr (2009) indicated that Cα of MSW with different degrees of biodegradation and values of maximum particle size were similar. Bareither et al. (2013) found that Cα of fresh and decomposed MSW was not affected by stress and specimen size. A relatively narrow range of about 0.01–0.05 for Cα was reported for MSW from these published studies. Many long-term laboratory and field studies on MSW settlement were carried out to understand biodegradationinduced compression. More satisfactory simulation for measured biodegradation-induced compression has been found on the basis of first-order biodegradation kinetics (Ivanova et al., 2008; Elagroudy et al., 2008; Gourc et al., 2010). The potential of biodegradation-induced compression strain was reported in a range of 0.13–0.27 for fresh MSW (Chen et al., 2010), and was believed to increase with increasing content of degradable matter. However, the determination of the rate of biodegradation-induced compression is still a challenge for researchers and engineers. Ivanova et al. (2008), Bareither et al. (2013) and Siddiqui et al. (2013) conducted pioneer studies on the relationship between the degrees of biodegradation-induced compression and biodegradation for MSW. Chen et al. (2010) proposed a one-dimensional (1-D) compression model which took two important compression characteristics of MSW into account: (i) the decrease of C’c due to biodegradation and (ii) the load dependence of secondary compression. This compression model was later applied by Chen et al. (2012) and Li et al. (2013) to analyze settlement and storage capacity of MSW landfills. However, in the model of Chen et al. (2010), creep and biodegradation-induced secondary compression were combined and expressed by first-order kinetics. In order to take a closer investigation of the relationship between biodegradation-induced compression and biodegradation process, a modified model separating out the effect of creep on secondary compression was proposed based on the model of Chen et al. (2010). Furthermore, the methods to determine the model parameters were proposed in this paper to make the modified model more applicable for engineers. Long-term laboratory compression experiments were carried out to test the proposed methods. Through simulation of compression data measured from the laboratory experiments as
t ε sc = Cα log10 tp
(3)
ε sb = ε sb∞ 1 − exp − c (t −tb ) (4)
where: εp represents the primary compression strain of waste responding to the surcharge load of σ (kPa); εsc and εsb are the creep and biodegradation-induced compression strains at the time t since the application of σ, respectively; σ0 (kPa) represents the pre-consolidation pressure; C’C0 represents the modified primary compression index for fresh waste; Cα is the creep index for fresh waste; tp (day) is the time for the completion of primary compression; εsb∞ is the ultimate biodegradation-induced secondary compression strain of the fresh waste under the applied load of σ; tb (day) is the time for the start of biodegradation; and c (day-1) is the rate coefficient of biodegradation-induced compression. The working principle of the modified model can be explained by a series of iso-stress compression curves (Figure 1) which represent the time-dependent compression under a constant stress. The real “loading” path of a specific waste layer within a landfill generally follows the path with respect to the applied load and elapsed time, e.g., OA1B1B2C2C3···Xi-2Xi-1Yi-1YiZi. Four assumptions are made including: (i) biodegradation rate of MSW is independent of the stress level; (ii) creep is independent of the stress level and the degree of biodegradation; (iii) the time for the completion of primary compression under a specific stress is short; and (iv) primary compressibility depends on the degree of biodegradation. Based on these assumptions, the total strain ε(σi,tj) at the point Zi is insensitive to the alternative “loading” paths, but depends only on the final applied load (i.e. σi) and the degree of biodegradation (i.e. the elapsed time tj). Therefore, ε(σi,tj) can be calculated by following the two-step “loading” path from point O to point Ai and then to point Zi. As shown in Figure 1, the “loading” path from point O to point Ai (i.e. load from zero to σi) is a monotonic loading applied on the fresh waste deposited. The subsequent path from point Ai to point Zi is
Downloaded from wmr.sagepub.com at The University of Iowa Libraries on June 5, 2015
201
Xu et al. 10 σ0
1
102
103
σ (kPa)
0 εsb∞(σ0) 20
εp(σ)
εsc(tĄ)
40
εsb∞(σ)
60 T0 T∞-sb T∞
ε (%)
Figure 2. Determination of parameters in the compression model.
a monotonic path with respect to time from zero to tj under the constant load of σi.
Determination of model parameters In Equation (2), two parameters including C’C0 and σ0 need to be determined. As shown in Figure 2, the value of C’C0 can be determined by the primary compression curve (i.e. line T0) obtained through laboratory 1-D mechanical compression tests on fresh waste. It is assumed that primary compression strain is zero if the applied load is smaller than σ0. Therefore, the stress at the intersection point of the primary compression curve for fresh waste and the zero strain coordinate is defined as σ0. In Equation (3), two parameters including Cα and tp need to be determined. The value of Cα can be determined from long-term laboratory compression tests on fresh waste under certain surcharge load with biodegradation inhibited. The time with respect to the inflexion point of the compression strain and logarithmic time curve curve for determining Cα is defined as tp, which means long-term compression strain becomes linear with logarithm of time after tp. In Equation (4), three parameters including εsb∞, tb and c need to be determined. A fully decomposed waste can be achieved by performing an accelerated biodegradation test on fresh waste under a sustained load (e.g. σ0). For a given stress σ, as shown in Figure 2, εsb∞(σ) can be determined from the vertical distance between the primary compression curves for fresh waste (i.e. line T0) and fully decomposed waste (i.e. line T∞) minus the creepinduced compression strain (i.e. εsc(t’) = Cαlog10(t’/tp)) caused during the test for achieving the fully decomposed waste. As compared with the time required for achieving the fully decomposed waste, the time required for primary compression test on the fully decomposed waste is assumed to be negligible. Therefore, line T∞-sb determining εsb∞(σ) is parallel to line T∞. εsb∞(σ) can be expressed as follows which indicates that biodegradation-induced compression is dependent on stress level:
σ ε sb∞ ( σ ) = (C ′C∞ − C ′C 0 )log10 + ε sb∞ ( σ0 ) σ0 t′ σ = (C ′C∞ − C ′C 0 )log10 + ε s 0 − Cα log10 σ0 tp
(5)
where: C’C∞ represents the modified primary compression index for the fully decomposed waste; t’ (day) represents the time required for the above accelerated biodegradation test for achieving the fully decomposed waste; εs0 is the total secondary compression strain of the accelerated biodegradation test on the fresh waste under a sustained load of σ0. However, it might be difficult to obtain a fully decomposed waste. The sample with the content of biodegradable matter less than 5% of the initial value of fresh waste is considered to have little variation of C’c after further degradation of the biodegradable matter left. This sample could be used to represent the fully decomposed waste. For a specific landfill, tb can be defined as the time when gas generation starts. The value of c can be obtained from the backanalysis of monitored long-term settlement data at landfills with similar decomposition conditions. Equation (4) can be rewritten as follows based on which the value of c can be determined by plotting ln(εsb∞(σ)- εsb(σ,t)) versus time and fitting the linear curve:
ln ε sb∞ (σ) − ε sb (σ) = −c(t − tb ) + ln ε sb∞ (σ) (6)
Materials and methods Experimental setup A schematic of the 1-D compression apparatus is shown in Figure 3. The specimen cell was made of a stainless steel cylinder with the inner diameter of 200 mm, wall thickness of 10 mm and height of 600 mm. A filter plate was placed at the bottom of the cell and a geotextile was placed above the plate. A steel plate with dead weight placed above was used to produce the compression load on top of the waste specimen. Settlement of the specimen was monitored by a LVDT (5000HCD-200, Weihe Ltd). Two O-rings were set around the load distribution plate and lubricated with vaseline oil to create gas tight seals and minimize the friction between the cell and the load distribution plate. Leachate was collected from the base of the specimen cell. The collected leachate was mixed with sewage sludge and recirculated into the waste specimen through the peristaltic pump in order to enhance biodegradation. The sewage sludge was obtained from the anaerobic digester of a municipal sewage plant in Hangzhou, China. Anaerobic digestion bacteria (e.g. methanogenic bacteria) contained in the sewage sludge could enhance the anaerobic degradation of fresh MSW which generally has little anaerobic digestive bacteria initially (Ivanova, 2007). A porous plate at the bottom of the load distribution plate was used to create uniform distribution of recirculated leachate on the specimen. A gas pressure gauge and a gas collection tube were installed at the top of the load distribution plate. The volume of generated gas was measured by means of the water (i.e. NaHCO3 saturated water) displacement technique. The temperature of the waste specimen was controlled by the means of water bath. A calefaction tube and a cyclic water pump were installed at the bottom of the water tank which
Downloaded from wmr.sagepub.com at The University of Iowa Libraries on June 5, 2015
202
Waste Management & Research 33(2)
Figure 3. Schematic of the compression apparatus. Table 1. Composition of MSW. Composition
Food
Paper
Wood
Textile
Plastic
Metal Glass
Cinder
Others*
(kg kg-1, %, wet basis)
67.7
6.7
0.6
0.6
8.5
1.5
13.7
0.7
*Undistinguishable soil-like matters.
was heat-insulated. The cyclic water pump was used to create a uniform temperature within the water bath. Hartz et al. (1982) found that 41°C was the optimum for the production of methane, and methane production would cease somewhere between 48°C and 55°C. Therefore, a threshold temperature of 41±1°C was set for the calefaction tube to produce optimal temperature condition for waste biodegradation.
obtain a waste specimen with the pre-designed void ratio and unit weight. The predesigned values of void ratio (e = 3.0) and unit weight (γ = 12.17 kN m-3) of the specimen are close to the reported data of waste drilled from the shallow layer of Qizishan landfill, China, respectively (Chen et al., 2009).
Tested materials
Three compression tests were carried out in cell 1, cell 2 and cell 3, respectively. The initial heights of specimens in cells 1, 2 and 3 were 400, 200 and 400 mm, respectively. The amount of waste required for each cell was determined by the pre-designed initial height and the unit weight mentioned above. In order to determine the parameters of the compression model through the proposed methods, each compression test consisted of three stages in time scale: (i) short-term multi-stage compression on fresh waste until reaching the pre-designed sustained load; (ii) continuing long-term compression under the predesigned sustained load with biodegradation enhanced; and (iii) final short-term multi-stage compression on the decomposed waste after secondary compression nearly stopped under the pre-designed sustained load. The measured primary compression strain in stage (i) and (iii) was used to determine C’C0, σ0, C’C∞
Synthetic MSW was made according to the reported composition of fresh MSW generated in Hangzhou, China. As shown in Table 1, degradable matter in waste is mainly composed of food waste (about 67.7%, wet basis) which has a water content of about 85% (He, 2011). Two subsamples of synthetic MSW were used for the determination of physical properties. Waste materials were dried at an oven temperature of 60°C for 48 hours to prevent ignition loss of organic matters, and the water content of the entire specimen (w = 54.6%, wet basis) was measured. The bulk specific gravity (ds = 2.21) of waste was measured using the water replacement method in accordance with the technical code of CJJ176-2012 (MOHURD, 2012). A certain amount of shredded waste (maximum particle size
