Waste Management xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Waste Management journal homepage: www.elsevier.com/locate/wasman
Thermogravimetric characteristics of typical municipal solid waste fractions during co-pyrolysis Hui Zhou, YanQiu Long, AiHong Meng, QingHai Li, YanGuo Zhang ⇑ Key Laboratory for Thermal Science and Power Engineering of Ministry of Education, Department of Thermal Engineering, Tsinghua University, Beijing 100084, PR China
a r t i c l e
i n f o
Article history: Received 28 July 2014 Accepted 29 September 2014 Available online xxxx Keywords: TG Interaction Municipal solid waste Pyrolysis PVC
a b s t r a c t The interactions of nine typical municipal solid waste (MSW) fractions during pyrolysis were investigated using the thermogravimetric analyzer (TGA). To compare the mixture results with the calculation results of superposition of single fractions quantitatively, TG overlap ratio was introduced. There were strong interactions between orange peel and rice (overlap ratio 0.9736), and rice and poplar wood (overlap ratio 0.9774). The interactions of mixture experiments postponed the peak and lowered the peak value. Intense interactions between PVC and rice, poplar wood, tissue paper, wool, terylene, and rubber powder during co-pyrolysis were observed, and the pyrolysis at low temperature was usually promoted. The residue yield was increased when PVC was blended with rice, poplar wood, tissue paper, or rubber powder; while the residue yield was decreased when PVC was blended with wool. Ó 2014 Elsevier Ltd. All rights reserved.
1. Introduction The amounts of municipal solid waste (MSW) are increasing rapidly (He et al., 2010). Traditional landfill method is facing many problems, such as land occupation, underground water pollution, and air pollution (Lee et al., 2014; Dong et al., 2003). MSW incineration has some advantages, such as volume reduction and energy recovery (Huai et al., 2008; Zhang et al., 2012). However, it also causes problems such as the release of dioxins (Zainal et al., 2014; Bogdal et al., 2013). In recent years, MSW pyrolysis is getting more and more concerns, because it may ease the problem of dioxins pollution and produce char, oil and syngas that can be further utilized (Luo et al., 2010; López et al., 2010). MSW is a very complex mixture, and the composition of MSW varies from place to place and time to time. Therefore, research of MSW mixture is only meaningful for specific fractions of MSW, which means the experiments can be hardly repeated (Zhou et al., 2014). For this reason, increasing amounts of research began to focus on single fractions of MSW. For example, Luo et al. (2010) pyrolyzed three representative fractions (plastic, kitchen garbage and wood) in a fixed bed reactor to evaluate the influence of particle size on pyrolysis performance of single-component MSW. Zheng et al. (2009) studied the pyrolysis characteristics of six representative organic fractions of MSW (wood chips, fabric, food residue, rubber, PE and wastepaper) in a specially designed ⇑ Corresponding author. Tel.: +86 10 62783373; fax: +86 10 62798047x801. E-mail address: [email protected] (Y. Zhang).
thermogravimetric analysis (TGA) apparatus with a maximum heating rate of 864.8 °C min1. Li et al. (1999) tested eight kinds of MSW fractions (paper, paperboard, waste plastics including PVC and PE, rubber, vegetal materials, wood, and orange husk) in a lab-scale rotary-kiln pyrolyser. However, most of the research studies the characteristics of single fractions, while the fractions do not act independently during pyrolysis (Wu et al., 2014). Sorum et al. (2001) investigated possible interactions between different paper and plastic fractions, and it was found that the reactivity of cellulosic matter was increased in a mixture with PVC. McGhee et al. (1995) pyrolyzed the mixtures of PVC with straw, and reported that the char yields were greater than produced by pyrolysis of the individual fractions due to the interaction of HCl and cellulose below 600 K. Zheng et al. (2009) suggested that the presence of PE weakened the reaction intensity of biomass component during fast pyrolysis. This could be attributed to the endothermic reaction of PE pyrolysis to decrease the heating rate of the feedstock. However, when the heating rate was lower, its decrease caused by PE was comparatively lower. Most of researches focused on the binary interaction of two or three fractions, the interactions between various kinds of fractions have not been studied systematically. The mean physical composition of MSW in Chinese cities is shown in Fig. 1. In combustible MSW, the contents of food residue, plastics, paper, textiles, wood waste and rubber, in decreasing order, were 55.86%, 11.15%, 8.52%, 3.16%, 2.94% and 0.84% (Zhou et al., 2014).
http://dx.doi.org/10.1016/j.wasman.2014.09.027 0956-053X/Ó 2014 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Zhou, H., et al. Thermogravimetric characteristics of typical municipal solid waste fractions during co-pyrolysis. Waste Management (2015), http://dx.doi.org/10.1016/j.wasman.2014.09.027
2
H. Zhou et al. / Waste Management xxx (2015) xxx–xxx
2. Material and methods 2.1. Materials
Fig. 1. The mean physical composition of MSW in China.
In this paper, nine kinds of MSW typical fractions were selected to study the thermogravimetric characteristics of interactions. According to the MSW fractions as reported in Fig. 1, orange peel and rice are representatives of food residue; poplar wood is representative of wood waste; tissue paper is representative of paper; wool and terylene are representatives of textiles; PE and PVC are representatives of plastics; and rubber powder was representative of rubber. The results of mixture experiments were compared with the calculation results of superposition of single fractions. Overlap ratio was applied to evaluate the interactions quantitatively.
The MSW fractions included orange peel, rice, poplar wood, tissue paper, wool, terylene, PE, PVC and rubber powder. Orange peel, rice and tissue paper were collected from the supermarket or restaurant in Tsinghua University, and poplar wood was collected from trees in Tsinghua University. Wool and terylene were collected from a shopping mall in Beijing, PE and PVC were provided by Shanghai Yangli Mechanical and Electrical Technology Co., Ltd. Rubber powder was provided by Hangzhou Boyang Rubber Chemical Co., Ltd. The proximate and ultimate analyses of the samples were shown in Table 1. Poplar wood and rubber powder have relatively high ash content, plastics (PE and PVC) show the highest volatile, orange peel and rubber powder have the highest fixed carbon content. The highest volatile was also reported by other research (Sorum et al., 2001). PE and rubber powder have the highest elemental carbon content (more than 80%), and they have little oxygen content, as reported by Sorum et al. (2001). PE has the highest hydrogen content of 11.20%, which has also been reported by other research (Li et al., 1999). The nitrogen and sulfur content of all the samples are less than 2% with exception of wool. PVC has a chlorine content of 56.35%, which was also reported by Li et al. (Li et al., 1999). The high heating value (HHV) of the samples except PE and rubber powder varies from 17.25 to 20.92 MJ kg1, while
Table 1 Proximate and ultimate analyses of MSW fractions. MSW classification
Samples
Food residue
Orange peel Rice Poplar wood Tissue paper Wool Terylene PE PVC Rubber powder
Wood waste Paper Textiles Plastics Rubber
Proximate analysis
HHVd (MJ kg1)
Ultimate analysis
Ad%
Vd%
FCd%
Cd%
Hd%
Od%
Nd%
St,d%
2.91 0.40 7.54 0.52 1.24 0.49 0.00 0.00 10.24
76.49 84.42 73.85 90.47 84.76 88.60 99.98 94.93 62.83
20.60 15.18 18.61 9.01 14.00 10.91 0.02 5.07 26.93
47.32 45.79 47.49 44.95 59.33 61.86 85.98 38.34 80.36
5.75 6.32 5.45 6.10 4.19 4.12 11.20 4.47 6.01
42.45 45.56 37.91 48.07 31.09 32.96 2.44 56.35a 0.96
1.39 1.68 1.41 0.25 2.62 0.29 0.21 0.23 0.62
0.18 0.25 0.20 0.11 1.53 0.28 0.17 0.61 1.81
18.47 18.14 18.50 17.25 20.92 20.86 46.48 20.83 35.74
A: ash; V: volatile; FC: fixed carbon; HHV: high heating value; d: dry basis. a It is Cl for PVC.
(a)
(b)
Fig. 2. TG curves of MSW fractions pyrolysis at a heating rate of 10 °C min1.
Please cite this article in press as: Zhou, H., et al. Thermogravimetric characteristics of typical municipal solid waste fractions during co-pyrolysis. Waste Management (2015), http://dx.doi.org/10.1016/j.wasman.2014.09.027
3
H. Zhou et al. / Waste Management xxx (2015) xxx–xxx
(a)
(b)
Fig. 3. DTG curves of MSW fractions pyrolysis at a heating rate of 10 °C min1.
the PE has the highest HHV of 46.48 MJ kg1, as reported by Zheng et al. (2009), followed by rubber powder (35.74 MJ kg1). Before pyrolysis, the samples were dried at 105 °C to eliminate the moisture. Samples were ground to less than 500 lm, which is small enough to prevent heat transfer effect in isothermal and dynamic experiments (Bockhorn et al., 1999). 2.2. Experimental apparatus The TGA experiments were performed by a NETZSCH STA 409C/ 3/F with a flow rate of 100 ml min1 of N2. Temperature rose from room temperature to 1000 °C at a heating rate of 10 °C min1. Repeated experiments showed that TG curves had good reproducibility. 3. Results and discussion 3.1. Pyrolysis of single fractions The TG and derivative thermogravimetric (DTG) curves of single fractions are shown in Figs. 2 and 3, respectively. As shown in Fig. 2(a), the residue of tissue paper was the lowest, which was consistent with the low fixed carbon and ash content in Table 1. Fig. 2(a) shows that the pyrolysis of orange peel began very early at approximate 200 °C. As shown in Fig. 3(a), the DTG curve of orange peel has two main peaks. The first one is at 226 °C, because of the decomposition of pectin and hemi-cellulose
(a)
(Fisher et al., 2002). The second one is at 333 °C, due to the decomposition of cellulose (Zhou et al., 2013). The DTG curve of rice and tissue paper only has one main peak, as shown in Fig. 3(a). The peak of rice was at 305 °C, due to the decomposition of starch, the main component of rice (Dumitriu, 2004). The peak of tissue paper was at 353 °C, due to the decomposition of cellulose. It should be noticed that tissue paper had obvious mass loss from 370 to 875 °C, because of the lignin content in tissue paper (Sorum et al., 2001). The pyrolysis of poplar wood has several peaks, as shown in Fig. 3(a). The first peak at 280 °C was a shoulder peak, because of the pyrolysis of hemi-cellulose. The second peak at 345 °C was the main peak, because of the pyrolysis of cellulose (Zhou et al., 2013). The mass loss after that without obvious peak was due to the slow decomposition of lignin in poplar wood. As shown in Figs. 2(b) and 3(b), PVC begins to lose mass at an early time (260 °C), because of the dehydrochlorination process. The second stage, which is attributed to the degradation of remaining hydrocarbons (Sorum et al., 2001), occurs from 405 to 535 °C. The pyrolysis of wool shows the maximum mass loss rate at 404 °C, because of the decomposition of keratin (Aluigi et al., 2014). The DTG curve of terylene has the peak at 436 °C. The peak of PE pyrolysis is quite high, at 479 °C. The pyrolysis of rubber has the main peak at 378 °C. The residue showed by TG curves, as reported in Fig. 2, follows the decreasing order rubber > wool > terylene > PVC > PE, which was congruent with the proximate results, as shown in Table 1.
(b)
R=0.9352
Fig. 4. Illustration of TG overlap ratio (a, poplar wood and wool; b, tissue paper and PVC).
Please cite this article in press as: Zhou, H., et al. Thermogravimetric characteristics of typical municipal solid waste fractions during co-pyrolysis. Waste Management (2015), http://dx.doi.org/10.1016/j.wasman.2014.09.027
4
H. Zhou et al. / Waste Management xxx (2015) xxx–xxx
Table 2 TG overlap ratio of pyrolysis of binary mixtures. Orange peel Orange peel Rice Poplar wood
0.9736 0.9900
Tissue paper Wool Terylene PE PVC
0.9936 0.9927 0.9899 0.9898 0.9852
Rubber powder
0.9842
Rice
Poplar wood
Tissue paper
Wool
Terylene
0.9774 0.9852 0.9888 0.9931 0.9922
0.9931 0.9936 0.9856 0.9872
0.9889 0.9822 0.9809
0.9874 0.9882
0.9872
0.9609 0.9850
0.9775 0.9871
0.9352 0.9875
0.9575 0.9873
0.9771 0.9890
PE
PVC
Rubber powder
0.9870 0.9873
0.9498
(b) rice+poplar wood R=0.9774
(a) rice+orange peel R=0.9736
Fig. 5. TG characteristics of interactions of rice with orange peel and poplar wood.
(a) rice+orange peel
(b) rice+poplar wood
Fig. 6. DTG characteristics of interactions of rice with orange peel and poplar wood.
3.2. Pyrolysis of binary mixtures 3.2.1. TG overlap ratio of binary mixtures To investigate the interaction of fractions during pyrolysis, every two fractions were mixed by weight at the ratio 1:1. If the two fractions pyrolyzed independently, and there was no interaction among them, the calculated TG curves of the synthetic samples would be plotted by sum of each fraction’s mass losses multiplying its proportion. The interactions of each MSW fractions during pyrolysis could be observed by comparing the experimental and calculated TG and DTG curves.
To evaluate the influence of integration on thermogravity, overlap ratio of TG curves was introduced (Zhou et al., 2013), which could be defined as:
R¼1 ¼1
The area sandwiched by two curves The total area R te D mðtÞ j jdt ts ðt e t s Þðms me Þ
ð1Þ
In Eq. (1), |Dm(t)| denoted the absolute value of difference of two TG curves corresponding temperature, and ts, te, ms and me
Please cite this article in press as: Zhou, H., et al. Thermogravimetric characteristics of typical municipal solid waste fractions during co-pyrolysis. Waste Management (2015), http://dx.doi.org/10.1016/j.wasman.2014.09.027
5
H. Zhou et al. / Waste Management xxx (2015) xxx–xxx
(a) PVC+rice
(b) PVC+poplar wood
(c) PVC+tissue paper
R=0.9775
R=0.9609
(d) PVC+wool
R=0.9352
(e) PVC+terylene
R=0.9575
(f) PVC+rubber powder
R=0.9771
R=0.9498
Fig. 7. TG characteristics of interactions of PVC with rice, poplar wood, tissue paper, wool, terylene, and rubber powder.
(a) PVC+rice
(b) PVC+poplar wood
(d) PVC+wool
(c) PVC+tissue paper
(e) PVC+terylene
(f) PVC+rubber powder
Fig. 8. DTG characteristics of interactions of PVC with rice, poplar wood, tissue paper, wool, terylene, and rubber powder.
denoted initial temperature, end temperature, initial mass (100%) and residue mass, respectively. R = 1 when two curves coincide, and 0 6 R < 1 in other cases. N sets of data were exported from the TG curve and the TG overlap ratio was determined as follows. The ideal N should tend to be positive infinity, but according to the definition of definite integral, when N is large enough, the difference is very small. During the calculation, N was taken as 16,000, which was large enough to reduce the error.
X ðte t s Þ lim R¼1
N!1
X jDmðtÞj
jDmðtÞj
N N
ðte ts Þðms me Þ
¼ 1 lim
N!1
N
N
ð2Þ
An overlap ratio close to 1 means little interactions between two fractions. Lower overlap ratios means stronger interactions between two fractions. To explain the concept of overlap ratio visually, Fig. 4 gives two examples. The interaction of poplar wood and wool is insignificant, with the overlap ratio 0.9936, as shown
Please cite this article in press as: Zhou, H., et al. Thermogravimetric characteristics of typical municipal solid waste fractions during co-pyrolysis. Waste Management (2015), http://dx.doi.org/10.1016/j.wasman.2014.09.027
6
H. Zhou et al. / Waste Management xxx (2015) xxx–xxx
in Fig. 4(a). The two curves from mixture experiment and linear superposition are almost coincident. The interaction of tissue paper and PVC is strong, with the overlap ratio 0.9352, as shown in Fig. 4(b). It can be seen that the difference between two curves are quite large. The TG overlap ratios of the binary mixtures of nine fractions are shown in Table 2. The overlap ratios lower than 0.9800 are underlined. The interactions between PE and biomass fraction (orange peel, poplar wood, and tissue paper) were not significant, with the overlap ratios larger than 0.9800. These results are supported by Jakab et al. (2001), where the interactions of MSW fractions were investigated, and wood, cellulose and lignin have a small effect on the thermogravimetric curves of PE. There are strong interactions between different biomass, in particular between orange peel and rice, and rice and poplar wood, with values of overlap ratio equal to 0.9736 and 0.9774, respectively. It should be noticed that PVC has strong interactions with all the other fractions, except plastic polymers during co-pyrolysis in TGA. Particularly, the interactions between PVC and rice, tissue paper, wool, or rubber powder are stronger, with the overlap ratios lower than 0.9700. The interactions between PVC and tissue paper are the most significant, with the overlap ratio 0.9352. PVC had weak interaction with PE, and the reason might be that with a low heating rate (10 °C min1), the pyrolysis of PE started from approximate 400 °C, while the main process of PVC pyrolysis was ended at this temperature, as shown in Figs. 2(b) and 3(b). The pyrolysis of PVC and PE did not occur at the same time, thus the interactions between them was weak. This was proved by the result that with higher heating rate of 40 °C min1, the interaction between PVC and PE was more significant (Wu et al., 2014). 3.2.2. Analysis of significant interactions As shown in Table 2, rice has significant interactions with orange peel and poplar wood. The TG curves are shown in Fig. 5. The interactions between rice and orange peel are similar to that of rice and poplar wood. There are no interactions before 300 °C. Above 300 °C, the interactions promote the pyrolysis, as shown in Fig. 5. While at 1000 °C, the residue of mixture experiments is similar to that of superposition results. The comparison of DTG curves are shown in Fig. 6. For the mixture of rice and orange peel, compared to linear superposition results, the mixture experiment got lower peak value, and the peak occurred later. It was same for the interaction between rice and poplar wood. Meanwhile, the second peak of linear superposition of rice and poplar wood disappeared in mixture experiments. During the mixture experiments, PVC had significant interactions (overlap ratio
Contents lists available at ScienceDirect
Waste Management journal homepage: www.elsevier.com/locate/wasman
Thermogravimetric characteristics of typical municipal solid waste fractions during co-pyrolysis Hui Zhou, YanQiu Long, AiHong Meng, QingHai Li, YanGuo Zhang ⇑ Key Laboratory for Thermal Science and Power Engineering of Ministry of Education, Department of Thermal Engineering, Tsinghua University, Beijing 100084, PR China
a r t i c l e
i n f o
Article history: Received 28 July 2014 Accepted 29 September 2014 Available online xxxx Keywords: TG Interaction Municipal solid waste Pyrolysis PVC
a b s t r a c t The interactions of nine typical municipal solid waste (MSW) fractions during pyrolysis were investigated using the thermogravimetric analyzer (TGA). To compare the mixture results with the calculation results of superposition of single fractions quantitatively, TG overlap ratio was introduced. There were strong interactions between orange peel and rice (overlap ratio 0.9736), and rice and poplar wood (overlap ratio 0.9774). The interactions of mixture experiments postponed the peak and lowered the peak value. Intense interactions between PVC and rice, poplar wood, tissue paper, wool, terylene, and rubber powder during co-pyrolysis were observed, and the pyrolysis at low temperature was usually promoted. The residue yield was increased when PVC was blended with rice, poplar wood, tissue paper, or rubber powder; while the residue yield was decreased when PVC was blended with wool. Ó 2014 Elsevier Ltd. All rights reserved.
1. Introduction The amounts of municipal solid waste (MSW) are increasing rapidly (He et al., 2010). Traditional landfill method is facing many problems, such as land occupation, underground water pollution, and air pollution (Lee et al., 2014; Dong et al., 2003). MSW incineration has some advantages, such as volume reduction and energy recovery (Huai et al., 2008; Zhang et al., 2012). However, it also causes problems such as the release of dioxins (Zainal et al., 2014; Bogdal et al., 2013). In recent years, MSW pyrolysis is getting more and more concerns, because it may ease the problem of dioxins pollution and produce char, oil and syngas that can be further utilized (Luo et al., 2010; López et al., 2010). MSW is a very complex mixture, and the composition of MSW varies from place to place and time to time. Therefore, research of MSW mixture is only meaningful for specific fractions of MSW, which means the experiments can be hardly repeated (Zhou et al., 2014). For this reason, increasing amounts of research began to focus on single fractions of MSW. For example, Luo et al. (2010) pyrolyzed three representative fractions (plastic, kitchen garbage and wood) in a fixed bed reactor to evaluate the influence of particle size on pyrolysis performance of single-component MSW. Zheng et al. (2009) studied the pyrolysis characteristics of six representative organic fractions of MSW (wood chips, fabric, food residue, rubber, PE and wastepaper) in a specially designed ⇑ Corresponding author. Tel.: +86 10 62783373; fax: +86 10 62798047x801. E-mail address: [email protected] (Y. Zhang).
thermogravimetric analysis (TGA) apparatus with a maximum heating rate of 864.8 °C min1. Li et al. (1999) tested eight kinds of MSW fractions (paper, paperboard, waste plastics including PVC and PE, rubber, vegetal materials, wood, and orange husk) in a lab-scale rotary-kiln pyrolyser. However, most of the research studies the characteristics of single fractions, while the fractions do not act independently during pyrolysis (Wu et al., 2014). Sorum et al. (2001) investigated possible interactions between different paper and plastic fractions, and it was found that the reactivity of cellulosic matter was increased in a mixture with PVC. McGhee et al. (1995) pyrolyzed the mixtures of PVC with straw, and reported that the char yields were greater than produced by pyrolysis of the individual fractions due to the interaction of HCl and cellulose below 600 K. Zheng et al. (2009) suggested that the presence of PE weakened the reaction intensity of biomass component during fast pyrolysis. This could be attributed to the endothermic reaction of PE pyrolysis to decrease the heating rate of the feedstock. However, when the heating rate was lower, its decrease caused by PE was comparatively lower. Most of researches focused on the binary interaction of two or three fractions, the interactions between various kinds of fractions have not been studied systematically. The mean physical composition of MSW in Chinese cities is shown in Fig. 1. In combustible MSW, the contents of food residue, plastics, paper, textiles, wood waste and rubber, in decreasing order, were 55.86%, 11.15%, 8.52%, 3.16%, 2.94% and 0.84% (Zhou et al., 2014).
http://dx.doi.org/10.1016/j.wasman.2014.09.027 0956-053X/Ó 2014 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Zhou, H., et al. Thermogravimetric characteristics of typical municipal solid waste fractions during co-pyrolysis. Waste Management (2015), http://dx.doi.org/10.1016/j.wasman.2014.09.027
2
H. Zhou et al. / Waste Management xxx (2015) xxx–xxx
2. Material and methods 2.1. Materials
Fig. 1. The mean physical composition of MSW in China.
In this paper, nine kinds of MSW typical fractions were selected to study the thermogravimetric characteristics of interactions. According to the MSW fractions as reported in Fig. 1, orange peel and rice are representatives of food residue; poplar wood is representative of wood waste; tissue paper is representative of paper; wool and terylene are representatives of textiles; PE and PVC are representatives of plastics; and rubber powder was representative of rubber. The results of mixture experiments were compared with the calculation results of superposition of single fractions. Overlap ratio was applied to evaluate the interactions quantitatively.
The MSW fractions included orange peel, rice, poplar wood, tissue paper, wool, terylene, PE, PVC and rubber powder. Orange peel, rice and tissue paper were collected from the supermarket or restaurant in Tsinghua University, and poplar wood was collected from trees in Tsinghua University. Wool and terylene were collected from a shopping mall in Beijing, PE and PVC were provided by Shanghai Yangli Mechanical and Electrical Technology Co., Ltd. Rubber powder was provided by Hangzhou Boyang Rubber Chemical Co., Ltd. The proximate and ultimate analyses of the samples were shown in Table 1. Poplar wood and rubber powder have relatively high ash content, plastics (PE and PVC) show the highest volatile, orange peel and rubber powder have the highest fixed carbon content. The highest volatile was also reported by other research (Sorum et al., 2001). PE and rubber powder have the highest elemental carbon content (more than 80%), and they have little oxygen content, as reported by Sorum et al. (2001). PE has the highest hydrogen content of 11.20%, which has also been reported by other research (Li et al., 1999). The nitrogen and sulfur content of all the samples are less than 2% with exception of wool. PVC has a chlorine content of 56.35%, which was also reported by Li et al. (Li et al., 1999). The high heating value (HHV) of the samples except PE and rubber powder varies from 17.25 to 20.92 MJ kg1, while
Table 1 Proximate and ultimate analyses of MSW fractions. MSW classification
Samples
Food residue
Orange peel Rice Poplar wood Tissue paper Wool Terylene PE PVC Rubber powder
Wood waste Paper Textiles Plastics Rubber
Proximate analysis
HHVd (MJ kg1)
Ultimate analysis
Ad%
Vd%
FCd%
Cd%
Hd%
Od%
Nd%
St,d%
2.91 0.40 7.54 0.52 1.24 0.49 0.00 0.00 10.24
76.49 84.42 73.85 90.47 84.76 88.60 99.98 94.93 62.83
20.60 15.18 18.61 9.01 14.00 10.91 0.02 5.07 26.93
47.32 45.79 47.49 44.95 59.33 61.86 85.98 38.34 80.36
5.75 6.32 5.45 6.10 4.19 4.12 11.20 4.47 6.01
42.45 45.56 37.91 48.07 31.09 32.96 2.44 56.35a 0.96
1.39 1.68 1.41 0.25 2.62 0.29 0.21 0.23 0.62
0.18 0.25 0.20 0.11 1.53 0.28 0.17 0.61 1.81
18.47 18.14 18.50 17.25 20.92 20.86 46.48 20.83 35.74
A: ash; V: volatile; FC: fixed carbon; HHV: high heating value; d: dry basis. a It is Cl for PVC.
(a)
(b)
Fig. 2. TG curves of MSW fractions pyrolysis at a heating rate of 10 °C min1.
Please cite this article in press as: Zhou, H., et al. Thermogravimetric characteristics of typical municipal solid waste fractions during co-pyrolysis. Waste Management (2015), http://dx.doi.org/10.1016/j.wasman.2014.09.027
3
H. Zhou et al. / Waste Management xxx (2015) xxx–xxx
(a)
(b)
Fig. 3. DTG curves of MSW fractions pyrolysis at a heating rate of 10 °C min1.
the PE has the highest HHV of 46.48 MJ kg1, as reported by Zheng et al. (2009), followed by rubber powder (35.74 MJ kg1). Before pyrolysis, the samples were dried at 105 °C to eliminate the moisture. Samples were ground to less than 500 lm, which is small enough to prevent heat transfer effect in isothermal and dynamic experiments (Bockhorn et al., 1999). 2.2. Experimental apparatus The TGA experiments were performed by a NETZSCH STA 409C/ 3/F with a flow rate of 100 ml min1 of N2. Temperature rose from room temperature to 1000 °C at a heating rate of 10 °C min1. Repeated experiments showed that TG curves had good reproducibility. 3. Results and discussion 3.1. Pyrolysis of single fractions The TG and derivative thermogravimetric (DTG) curves of single fractions are shown in Figs. 2 and 3, respectively. As shown in Fig. 2(a), the residue of tissue paper was the lowest, which was consistent with the low fixed carbon and ash content in Table 1. Fig. 2(a) shows that the pyrolysis of orange peel began very early at approximate 200 °C. As shown in Fig. 3(a), the DTG curve of orange peel has two main peaks. The first one is at 226 °C, because of the decomposition of pectin and hemi-cellulose
(a)
(Fisher et al., 2002). The second one is at 333 °C, due to the decomposition of cellulose (Zhou et al., 2013). The DTG curve of rice and tissue paper only has one main peak, as shown in Fig. 3(a). The peak of rice was at 305 °C, due to the decomposition of starch, the main component of rice (Dumitriu, 2004). The peak of tissue paper was at 353 °C, due to the decomposition of cellulose. It should be noticed that tissue paper had obvious mass loss from 370 to 875 °C, because of the lignin content in tissue paper (Sorum et al., 2001). The pyrolysis of poplar wood has several peaks, as shown in Fig. 3(a). The first peak at 280 °C was a shoulder peak, because of the pyrolysis of hemi-cellulose. The second peak at 345 °C was the main peak, because of the pyrolysis of cellulose (Zhou et al., 2013). The mass loss after that without obvious peak was due to the slow decomposition of lignin in poplar wood. As shown in Figs. 2(b) and 3(b), PVC begins to lose mass at an early time (260 °C), because of the dehydrochlorination process. The second stage, which is attributed to the degradation of remaining hydrocarbons (Sorum et al., 2001), occurs from 405 to 535 °C. The pyrolysis of wool shows the maximum mass loss rate at 404 °C, because of the decomposition of keratin (Aluigi et al., 2014). The DTG curve of terylene has the peak at 436 °C. The peak of PE pyrolysis is quite high, at 479 °C. The pyrolysis of rubber has the main peak at 378 °C. The residue showed by TG curves, as reported in Fig. 2, follows the decreasing order rubber > wool > terylene > PVC > PE, which was congruent with the proximate results, as shown in Table 1.
(b)
R=0.9352
Fig. 4. Illustration of TG overlap ratio (a, poplar wood and wool; b, tissue paper and PVC).
Please cite this article in press as: Zhou, H., et al. Thermogravimetric characteristics of typical municipal solid waste fractions during co-pyrolysis. Waste Management (2015), http://dx.doi.org/10.1016/j.wasman.2014.09.027
4
H. Zhou et al. / Waste Management xxx (2015) xxx–xxx
Table 2 TG overlap ratio of pyrolysis of binary mixtures. Orange peel Orange peel Rice Poplar wood
0.9736 0.9900
Tissue paper Wool Terylene PE PVC
0.9936 0.9927 0.9899 0.9898 0.9852
Rubber powder
0.9842
Rice
Poplar wood
Tissue paper
Wool
Terylene
0.9774 0.9852 0.9888 0.9931 0.9922
0.9931 0.9936 0.9856 0.9872
0.9889 0.9822 0.9809
0.9874 0.9882
0.9872
0.9609 0.9850
0.9775 0.9871
0.9352 0.9875
0.9575 0.9873
0.9771 0.9890
PE
PVC
Rubber powder
0.9870 0.9873
0.9498
(b) rice+poplar wood R=0.9774
(a) rice+orange peel R=0.9736
Fig. 5. TG characteristics of interactions of rice with orange peel and poplar wood.
(a) rice+orange peel
(b) rice+poplar wood
Fig. 6. DTG characteristics of interactions of rice with orange peel and poplar wood.
3.2. Pyrolysis of binary mixtures 3.2.1. TG overlap ratio of binary mixtures To investigate the interaction of fractions during pyrolysis, every two fractions were mixed by weight at the ratio 1:1. If the two fractions pyrolyzed independently, and there was no interaction among them, the calculated TG curves of the synthetic samples would be plotted by sum of each fraction’s mass losses multiplying its proportion. The interactions of each MSW fractions during pyrolysis could be observed by comparing the experimental and calculated TG and DTG curves.
To evaluate the influence of integration on thermogravity, overlap ratio of TG curves was introduced (Zhou et al., 2013), which could be defined as:
R¼1 ¼1
The area sandwiched by two curves The total area R te D mðtÞ j jdt ts ðt e t s Þðms me Þ
ð1Þ
In Eq. (1), |Dm(t)| denoted the absolute value of difference of two TG curves corresponding temperature, and ts, te, ms and me
Please cite this article in press as: Zhou, H., et al. Thermogravimetric characteristics of typical municipal solid waste fractions during co-pyrolysis. Waste Management (2015), http://dx.doi.org/10.1016/j.wasman.2014.09.027
5
H. Zhou et al. / Waste Management xxx (2015) xxx–xxx
(a) PVC+rice
(b) PVC+poplar wood
(c) PVC+tissue paper
R=0.9775
R=0.9609
(d) PVC+wool
R=0.9352
(e) PVC+terylene
R=0.9575
(f) PVC+rubber powder
R=0.9771
R=0.9498
Fig. 7. TG characteristics of interactions of PVC with rice, poplar wood, tissue paper, wool, terylene, and rubber powder.
(a) PVC+rice
(b) PVC+poplar wood
(d) PVC+wool
(c) PVC+tissue paper
(e) PVC+terylene
(f) PVC+rubber powder
Fig. 8. DTG characteristics of interactions of PVC with rice, poplar wood, tissue paper, wool, terylene, and rubber powder.
denoted initial temperature, end temperature, initial mass (100%) and residue mass, respectively. R = 1 when two curves coincide, and 0 6 R < 1 in other cases. N sets of data were exported from the TG curve and the TG overlap ratio was determined as follows. The ideal N should tend to be positive infinity, but according to the definition of definite integral, when N is large enough, the difference is very small. During the calculation, N was taken as 16,000, which was large enough to reduce the error.
X ðte t s Þ lim R¼1
N!1
X jDmðtÞj
jDmðtÞj
N N
ðte ts Þðms me Þ
¼ 1 lim
N!1
N
N
ð2Þ
An overlap ratio close to 1 means little interactions between two fractions. Lower overlap ratios means stronger interactions between two fractions. To explain the concept of overlap ratio visually, Fig. 4 gives two examples. The interaction of poplar wood and wool is insignificant, with the overlap ratio 0.9936, as shown
Please cite this article in press as: Zhou, H., et al. Thermogravimetric characteristics of typical municipal solid waste fractions during co-pyrolysis. Waste Management (2015), http://dx.doi.org/10.1016/j.wasman.2014.09.027
6
H. Zhou et al. / Waste Management xxx (2015) xxx–xxx
in Fig. 4(a). The two curves from mixture experiment and linear superposition are almost coincident. The interaction of tissue paper and PVC is strong, with the overlap ratio 0.9352, as shown in Fig. 4(b). It can be seen that the difference between two curves are quite large. The TG overlap ratios of the binary mixtures of nine fractions are shown in Table 2. The overlap ratios lower than 0.9800 are underlined. The interactions between PE and biomass fraction (orange peel, poplar wood, and tissue paper) were not significant, with the overlap ratios larger than 0.9800. These results are supported by Jakab et al. (2001), where the interactions of MSW fractions were investigated, and wood, cellulose and lignin have a small effect on the thermogravimetric curves of PE. There are strong interactions between different biomass, in particular between orange peel and rice, and rice and poplar wood, with values of overlap ratio equal to 0.9736 and 0.9774, respectively. It should be noticed that PVC has strong interactions with all the other fractions, except plastic polymers during co-pyrolysis in TGA. Particularly, the interactions between PVC and rice, tissue paper, wool, or rubber powder are stronger, with the overlap ratios lower than 0.9700. The interactions between PVC and tissue paper are the most significant, with the overlap ratio 0.9352. PVC had weak interaction with PE, and the reason might be that with a low heating rate (10 °C min1), the pyrolysis of PE started from approximate 400 °C, while the main process of PVC pyrolysis was ended at this temperature, as shown in Figs. 2(b) and 3(b). The pyrolysis of PVC and PE did not occur at the same time, thus the interactions between them was weak. This was proved by the result that with higher heating rate of 40 °C min1, the interaction between PVC and PE was more significant (Wu et al., 2014). 3.2.2. Analysis of significant interactions As shown in Table 2, rice has significant interactions with orange peel and poplar wood. The TG curves are shown in Fig. 5. The interactions between rice and orange peel are similar to that of rice and poplar wood. There are no interactions before 300 °C. Above 300 °C, the interactions promote the pyrolysis, as shown in Fig. 5. While at 1000 °C, the residue of mixture experiments is similar to that of superposition results. The comparison of DTG curves are shown in Fig. 6. For the mixture of rice and orange peel, compared to linear superposition results, the mixture experiment got lower peak value, and the peak occurred later. It was same for the interaction between rice and poplar wood. Meanwhile, the second peak of linear superposition of rice and poplar wood disappeared in mixture experiments. During the mixture experiments, PVC had significant interactions (overlap ratio
