Kinetics and Mechanisms for Copyrolysis of Palm Empty Fruit Bunch

Subscriber access provided by UNIVERSITY OF CONNECTICUT

Article

Kinetics and mechanisms for co-pyrolysis of palm empty fruit bunch fibre (EFBF) with palm oil mill effluent (POME) sludge Yen Yee Chong, Suchithra Thangalazhy-Gopakumar, Suyin Gan, Hoon Kiat Ng, Lai-Yee Lee, and Sushil Adhikari Energy Fuels, Just Accepted Manuscript • DOI: 10.1021/acs.energyfuels.7b00877 • Publication Date (Web): 05 Jul 2017 Downloaded from http://pubs.acs.org on July 6, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Energy & Fuels is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 37

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 37

1

Kinetics and mechanisms for co-pyrolysis of palm

2

empty fruit bunch fibre (EFBF) with palm oil mill

3

effluent (POME) sludge

4

Yen Yee Chonga, Suchithra Thangalazhy-Gopakumara, *, Suyin Gana, Hoon Kiat Ngb, Lai Yee

5

Leea, Sushil Adhikaric a

6

Department of Chemical and Environmental Engineering, Faculty of Engineering,

7

University of Nottingham Malaysia Campus, Jalan Broga, Semenyih 43500, Selangor Darul

8

Ehsan, Malaysia

9

b

Department of Mechanical, Materials and Manufacturing Engineering, Faculty of

10

Engineering, University of Nottingham Malaysia Campus, Jalan Broga, Semenyih 43500,

11

Selangor Darul Ehsan, Malaysia

12

c

13 14 15 16 17 18 19

* E-mail: [email protected]; Tel: +6 (03) 8725 3635

20

Co-pyrolysis of biomass is one of the potential options to improve the quality of bio-oil. In

21

this study, different types of feedstock: palm empty fruit bunches fibre (EFBF) and palm oil

22

mill effluent (POME) sludge were performed via thermogravimetric analysis (TGA). The

23

thermogravimetric behaviour of EFBF and POME sludge blends (EFBF : POME sludge mass

24

ratio of 100%, 90%, 75%, 50%, 25%, and 0%) were subjected to different heating rate of (5,

Department of Biosystems Engineering, Auburn University, Auburn, AL 36849, USA

KEYWORDS. Co-pyrolysis, Thermogravimetric analysis (TGA), Synergic effect, Kinetics, Mechanisms ABTRACT

1 ACS Paragon Plus Environment

Page 3 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

25

10, 20, 30, 40 °C/min) with nitrogen (N2) purge of 20 ml/min to simulate pyrolysis

26

conditions. As the percentage of POME sludge in the blend increases, the thermo-gravimetric

27

data (TG) and thermo-gravimetric derivative (DTG) profiles shifted from EFBF to that of

28

POME sludge gradually. Higher mass loss rate of EFBF upon devolatilization indicates the

29

higher reactivity than that of POME sludge. During co-pyrolysis, a positive synergistic effect

30

was observed. All the samples experienced three pyrolysis stages and for each stage, the

31

mechanisms responsible were determined. Third order kinetic model (F3) was identified as

32

the most suitable model in master plot method. However, a deviation from theoretical master

33

plot at high percentage of POME sludge in blends was observed. Therfore, a stagewise

34

analysis of co-pyrolysis was done using Coats-Redfern (CR) method. A change in diffusion

35

mechanism was identified as POME sludge percentage increased in blends during the main

36

decomposition stage, which reveals the lack of specific shape for sludge particles.

37

1. Introduction

38

Concerns on the effects of fossil fuel emissions on the environment have been raised and as

39

fossil fuels are non-renewable energy sources, energy security is of another concern as well.

40

Malaysia is a rapidly developing country and is the third largest energy consumer in

41

Southeast Asia. In the last decade, Malaysia’s energy consumption grew at an average of

42

11.28 % 1. In an effort to encourage the employment of renewable resources, Malaysian

43

government announced renewable energy as the 5th fuel in the energy supply mix 2.

44

Among renewable energy sources, biomass shows high potential in terms of feed flexibility

45

and abundance. Malaysia is known as the largest exporter of palm oil, and currently accounts

46

for 39 % of world palm oil production 3. Noticeably large amount of palm waste is being

47

produced simultaneously. Shuit et al. (2009) have done a

48

sustainability of oil palm biomass in Malaysia 4. By taking the advantages of biomass energy

review on availability and


Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 37

49

production, studies on utilizing palm waste such as empty fruit bunch (EFB) and palm kernel

50

shells (PKS) have been flourishing.

51 52

Liquid fuel is the most consumed form of energy in the world 1. Thus, the finding of a

53

potential substitute or additive for liquid fuels (crude oil) is of importance. Fast pyrolysis, a

54

thermochemical conversion method of biomass to obtain liquid fuel, provides high liquid

55

yield and high energy density compared to virgin biomass 5. The liquid fuel obtained from

56

fast pyrolysis is known as bio-oil.

57 58

Generally, bio-oil derived from lignocellulosic biomass is acidic in nature, whereas bio-oil

59

derived from sludge is alkali in nature 5. Abdulla et al. (2011) characterized the bio-oil

60

derived from empty fruit bunch (EFB) and obtained a low pH of between 2 and 3, which was

61

contributed by the presence of organic acids, mostly acetic acid and formic acids 6. Having a

62

high acidity causes the bio-oil to be corrosive and thus, not suitable to be directly used as

63

fuel. On the contrary, the pH of the bio-oil derived from POME sludge has a high pH of

64

about 9.4, signifying that the bio-oil was alkaline in nature. Therefore, the blending of both

65

biomass and sludge derived bio-oils or co-pyrolysis of biomass and sludge in order to achieve

66

a neutralized bio-oil was suggested 5.

67 68

Co-pyrolysis of biomass is one of the potential options to improve the quality of bio-oil. Co-

69

pyrolysis of biomass is being widely explored by many researchers and has been reviewed by

70

Abnisa et al. (2014) 7. Due to the differences in the chemical and physical properties of

71

different biomass, a different thermal reactivity occurred during co-pyrolysis 8. A positive

72

synergistic effect was observed in previous co-pyrolysis studies , resulting in an increase in

73

both gas and liquid products, but decrease in char production 9–11. However, another study on


Page 5 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

74

co-pyrolysis of sewage sludge and pine sawdust did not show significant synergistic effect 12.

75

No significant effect was also observed for the co-pyrolysis of coal and oil palm biomass 13.

76

Apart from the mentioned synergistic effects, an inhibitive effect was observed in the co-

77

pyrolysis of petrochemical wastewater sludge with lignite 8.

78 79

Thermogravimetric analysis (TGA) is one of the most common methods to evaluate and

80

compare thermal kinetics during thermal conversion of biomass 14. Model free methods such

81

as Coats-Redfern (CR) have been broadly used to obtain the thermal kinetic triplets (apparent

82

reaction activation energy, E, apparent pre-exponential factor, A, and kinetic model, (α) by

83

applying the Arrhenius equation. To accurately predict and calculate the kinetics, Vyazovkin

84

took into consideration the fact that the reactions are non-isothermal and heterogeneous, and

85

proposed calculation methods that are currently being widely applied 15. Having known these

86

parameters, the degradation behaviour of the blends and their mechanisms can be further

87

understood, which can assist in the design and optimization of the operation 8.

88 89

The use of biomass as an energy source is indeed attractive due to the fact that biomass is

90

abundant. However, biomass is diverse in nature and exhibits different behaviours in thermal

91

process. Pyrolysis of EFB had been done by many and the pyrolysis of POME sludge had

92

been done by Thangalazhy-Gopakumar et al. (2015)

93

reactivity and kinetics of the co-pyrolysis of EFBF and POME sludge have yet to be done.

94

The current study investigated the pyrolytic behaviour for the co-pyrolysis of EFBF and

95

sludge with the aim to learn the interactions and influences between EFBF and sludge. This

96

study resulted in formulating the kinetics for the co-pyrolysis of EFBF and sludge using

97

TGA.

5,6,13

. However, the study on the

98


Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 37

99

The kinetic methods used to obtain the kinetic results were Vyazovkin and Coats-Redfern

100

(CR) methods. Vyaovkin method is an iterative method that assumed the independence of the

101

reaction model, () of the heating program, ( ). This method acknowledges the varying of

102

the activation energy, with the conversion degree, rather than assuming a constant at

103

different degrees of conversion. Vyazovkin method is a non-linear algorithm, that is able to

104

give an estimated activation energy with a mere 5% error 16,17. By using CR method, the TGA

105

curves of the samples were divided into three different stages and analysed individually: 1)

106

evaporation of moisture, 2) devolatilization of cellulose and hemicellulose, and 3)

107

decomposition of lignin. A depth analysis on the diffusion and reaction mechanisms and

108

their corresponding kinetic models in the co-pyrolysis of EFBF and sludge at different stages

109

was investigated.

110 111

2. Materials and Methods

112

2.1 Feedstock Characterization

113

The feedstocks used in this work were palm empty fruit bunches (EFBF) and treated palm oil

114

mill effluent (POME) sludge. The samples were collected from Seri Ulu Langat Palm Oil

115

Mill Sdn. Bhd, Dengkil, Selangor (Malaysia). EFBF was dried in an oven at 75 °C for 16 h.

116

Dried sludge, which was ready to compost in soil was collected and further sun dried for 3

117

days. The samples were then grounded to less than 2 mm in particle size. In order to better

118

understand the thermal conversion process of both fuels, ultimate and proximate analyses

119

were conducted and are presented in Table 1. The moisture content of the samples were

120

determined and reported using TGA results, where the weight % at around 150 °C was

121

deducted by the initial weight % of the sample, whereas volatile matter was determined from

122

the difference in the weight % at 150 °C and 900 °C. To further validate the moisture content

123

of the samples, the mass difference of both EFBF and POME sludge were calculated after 5 ACS Paragon Plus Environment

Page 7 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

124

being placed in the oven at 103 ˚C for 16 hours. The ash content of the sample was

125

determined per ASTM E 1755 standard. Next, fixed carbon was calculated by subtracting the

126

weight percentages of moisture content, volatile matter, and ash content from 100 %. Higher

127

heating values (HHV) of both EFBF and POME sludge were measured using Parr 6100 bomb

128

calorimeter. Different blends at different EFBF to POME sludge mass ratios were prepared

129

by mixing (100:0 90:10, 75:25, 50:50, 25:75, and 0:100). The analyses were triplicated to

130

ensure reproducibility of the results.

131

Table 1. Ultimate and proximate analyses and HHV of EFBF and POME sludge Sample

Proximate analysis

HHV (MJ/kg)

Moisture

Volatile

Ash

Fixed

content

matter

content

carbon

(%)

(%)

(%)

(%)

EFBF

5.3±0.3

74.0±1.0 1.3±0.2

19.4±0.8 17.6±0.9

POME

8.5±1.6

47.7±0.8 23.8±0.5

20.0±1.4 13.8±0.9

sludge 132

133

The volatile matter in sludge was very low as compared to that of EFBF, which was

134

expected. However the fixed carbon available in sludge was about the same as that of EFBF,

135

which showed the potential application of sludge bio-char from pyrolysis. As sludge has high

136

ash content, POME sludge was further examined by X-ray microanalysis, using FESEM –

137

Energy Dispersive X-ray Spectroscopy (EDX) in order to understand its inorganic content.

138

Table 2 shows the elements detected in POME sludge and the standards used. The elements

139

Ca, Mg, K, and Zn have studied for their catalytic activities for the thermal degradation of

140

biomass

18–21

. Presence of alkaline earth metals enhance the breakdown of biomass


Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 37

22

141

components, where the effect was increasing in the order of Mg, Ca, K, Na

142

oxygen content together with the presence of the mentioned elements in POME sludge

143

suggest the presence of the metal oxides in the sludge, which posed catalytic effect upon co-

144

pyrolysis of EFBF and sludge. Three points were taken for each microscopic analysis.

145

Table 2. Elements detected and the standard used in POME sludge from FESEM-EDX

146

analysis Element

Standard

C

CaCO3

50.4±4.0

O

SiO2

38.0±3.8

Mg

MgO

1.3±0.8

Al

Al2O3

0.7±0.1

Si

SiO2

1.2±0.3

P

GaP

2.1±0.8

S

FeS2

1.3±0.2

K

MAD-10 Feldspar

1.7±0.5

Ca

Wollastonite (CaSiO3)

1.3±0.2

Mn

Mn

1.7±0.9

. The high

Weight %

147

148

2.2 Thermogravimetric analysis (TGA)

149

Thermogravimetric analysis (TGA) was carried out in programmable TGA DSC 1 Mettler

150

Toledo to examine the decomposition behaviour of biomass upon pyrolysis. The different

151

EFBF to POME sludge ratios were taken for co-pyrolysis studies. The samples were heated at

152

heating rates of 5, 10, 20, 30 & 40 ˚C /min with nitrogen (N2) purge of 20 ml/min in order to

153

prevent loss of volatiles and to stimulate pyrolysis conditions. In each experimental run, 7 ACS Paragon Plus Environment

Page 9 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

154

approximately 10 mg of biomass was heated from room temperature to 900 ˚C, and held at

155

that temperature for 10 min. The experiments were duplicated to ensure the reproducibility of

156

results.

157

2.2.1 Synergic effect

158

To study the existence of interaction between the EFBF and POME sludge blends upon

159

pyrolysis, the theoretical values for thermo-gravimetric data (TG) and thermo-gravimetric

160

derivative (DTG) curves of the blends were calculated. Eqn. (1) was used to obtain the values

161

by adding the decomposition curves of each individual component 23:

162

= . + .

163

where, is the theoretical weight percentage (TG) or derivative of weight percentage (DTG)

164

of the blends; are experimental TG and DTG values of EFBF and POME,

165

respectively; are mass percentages of EFBF and POME in the blends.

166

(1)

2.2.2 Kinetic model

167

Kinetic analysis of EFBF, POME sludge, and their blends were carried out to obtain the

168

kinetic triplets (E, A, and ()).

169

As the pyrolysis of solid fuels is non-isothermal and heterogeneous, the pyrolysis process was

170

divided into innumerable isothermal stages by calculating the conversion degree for each

171

pyrolysis stage 8,15.

172

The conversion degree, is defined as the mass fraction of decomposed solid:

173

=

174

where !" , !, and !$ are the initial, instantaneous, and final masses of the solids,

175

respectively.

(2),


Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 37

176

In non-isothermal experiments, the rate of solid degradation can be written as:

177

%

= & ( )

(3),

178

where k is the temperature-dependant rate constant and () is a function of conversion that

179

varies according to the reaction model.

180

The reaction constant, &, can be expressed by Arrhenius equation

181

& = ' exp (−

182

where ' is the pre-exponential factor, is the activation energy, . is the gas constant, and

183

is the absolute temperature.

184

Substituting Eqn. (4) into Eqn. (3) gives

185

186

%

,-

)

(4),

= ' exp /− ,-0 . ( )

(5)

Considering a constant heating rate of 1 =

%

, Eqn. (5) can be rearranged to

187

188

Some methods employ the integral form of Eqn. (6), which can be presented in many forms,

189

as shown in Eqn. (7).

190

() = 45

191

where () is the integrated form of the conversion dependence function (), the

192

temperature integral 7() = 4< −(8 9 / ; ) , and = ,-.

193

-

2

-

= 3 exp /− ,-0 . ()

$()

2

-

(6)

2

= 3 4- exp / ,- 0 = /3,0 7() 6

9

(7),

2.2.2.1 Vyazovkin method


Page 11 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

194

Vyazovkin stated that in the case of using a linear heating program ( ) = 5 + 1 , the

195

integral, =( , ) which has no analytical solution is obtained:

196

() = 4- exp / 0 3 6 ,-

197

= /30 I(E, T)

198

By assuming that the reaction model is independent of the heating rate, Eqn (8) can be

199

written for a given conversion and a set of experiments performed under different heating

200

rates 1A (B = 1, … , ) as follows:

201

/ 3E 0 =G , ,H I = / 3E0 =G , ,; I = ⋯ = /3E 0 =G , ,M I

202

which leads to Eqn (10):

203

N( ) = O∑MAXH ∑MVWA Q[

204

where N( ) is the minimum of the function and activation energy at degree of conversion α,

205

, is determined as the value that minimizes the function, is the number of heating rates,

206

and =Y , ,A Z is the exponential integral, 7() that results from heating rate 1A . For the

207

approximation of 7() (Eqn. 7), nonlinear fourth degree Senum-Yang approximation (Eqn.

208

(11)) that was used as the approximation is highly accurate 24.

209

7( ) =

210

The detailed derivation of this advanced iso-conversional method is provided elsewhere 17,25.

211

-

2

2

2

(8)

2

F

2

J

Q[E ,-E,S ]3U

E ,-E,U ]3S

[\](9) 9

L

O

9 ^ _H`9 J _`a9_ba

. 9c _;59^ _H;59J _;d59_H;5

(9),

(10),

(11).

2.2.2.2 Master-plots Method

212

To identify the reaction model involved in the solid-state reaction, master plots method was

213

employed. Using a reference point of = 0.5, Eqn. (7) could then be represented as 10 ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

2

214

(0.5) = / 0 7(5.g )

215

Upon dividing Eqn. (7) by Eqn. (12), Eqn. (13) was obtained.

216

Page 12 of 37

(12)

3,

()

(5.g)

h(9)

= h(9

(13)

6.i )

()

217

Based on various () functions as presented in Table 3, theoretical master plots of

218

against were plotted. As for experimental master plots of

219

data obtained under any heating rates were used. Eqn. (13) indicates that when an appropriate

220

kinetic model is used, the values of (5.g) and h(9

()

h(9)

6.i )

h(9)

h(96.i )

(5.g)

against , experimental

would be equivalent at a given .

2.2.2.3 Coats-Redfern (CR) method

221

222

Besides that, the integral Coats-Redfern (CR) method was also used to evaluate and to

223

calculate the kinetic triplets for EFBF, POME sludge, and their blends. CR method is further

224

integrated, yielding:

225

ln l

226

As the temperature range applied in the combustion of the samples, the value of

227

less than one, so Eq. (15) is obtained.

228

ln l

229

where, ln(3) is essentially a constant value.

230

A straight line should be obtained from the plotting of ln l

231

linear relationship. If the correct () is used, the straight line should have a high correlation

() -J

2,

m = ln l

3

() -J

2,

/1 −

;,

0m −

(14)

,-

m = ln(3 ) − ,-

;,

was far

(15)

2,

() -J

m against

H

-

as they have a


Page 13 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

232

coefficient of linear regression analysis. Then, the values of E and A can be derived from the

233

slope − and the intercept ln( ), respectively 26.

234

By identifying () that gives the highest correlation coefficient, the pyrolysis reaction of

235

the samples can be associated with the appropriate mechanisms. The basic model functions

236

() that are used in this kinetic study of solid-state reactions are shown in the Table 3.

237

Table 3. Expressions of functions g (α) and their corresponding mechanisms (Adapted from

238

14,27

2,

,

No.

3

) Symbol

()

Name of

()

function

Ratedetermining mechanism

1. Chemical process or mechanism non-invoking equations 1.1

nH/o

One-third

(3/2)(1-α)

1/3

1-(1-α)

2/3

reaction

order 1.2

no/d

Three-quarters

4(1-α)

3/4

1-(1-α)

1/4

order 1.3

nH

Chemical

Chemical reaction

First order

1-α

Chemical

-ln(1-α)

reaction 1.4

no/;

One and a half

2(1-α)

3/2

(1-α)

-1/2

-1

order 1.5

n;

Second order

Chemical reaction

(1-α)

2

-1

(1-α) -1

Chemical reaction

1.6

no

Third order

(1/2)(1-α)

3

-2

(1-α) -1

Chemical reaction

2. Phase boundary reaction 2.1

.H , n5 , pH

Power law

(1-α)

0

α

Contracting


Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 37

disk 2.2

.; , nH/;

Power law

2(1-α)

1/2

1-(1-α)

1/2

Contracting cylinder

2.3

.o , n;/o

Power law

3(1-α)

2/3

1-(1-α)

1/3

Contracting sphere

3. Based on the diffusion mechanism 3.1

qH

Parabola low

α2

1/2α

Onedimensional diffusion

3.2

q;

Valensi

[- lnG1-αI ]

-1

α+G1-αIln(1-α)

equation

Twodimensional diffusion

3.3

qo

Jander

1/3 2

1/3 -1

2/3

[1-G1-αI ]

(3/2)(1-α) [1-(1-α) ]

equation

Threedimensional diffusion, spherical symmetry

3.4

qd

Ginstling(3/2)[G1-αI

-1/3

-1

1-2α/3-(1-α)

-1]

2/3

Three-

Brounstein

dimensional

equation

diffusion, cylindrical symmetry

3.5

qg

Zhuravlev,

4/3

1/3

-1

-1/3

[G1-αI

(3/2)(1-α) [G1-αI -1]

2

-1]

Three-

Lasokin,

dimensional

Tempelman

diffusion

equation 3.6

qa

anti-Jander

(3/2)(1+α)2/3 [(1+α)1/3 -1]

-1

[(1+α)1/3 -1]

2

Three-


Page 15 of 37

equation

dimensional diffusion

qr

3.7

q`

3.8

anti-Ginstling-

(3/2)[(1+α)-1/3 -1]

1+2α/3-(1+α)2/3

-1

Three-

Brounstein

dimensional

equation

diffusion

anti-

(3/2)(1+α)4/3 [(1+α)-1/3-1]

-1

[(1+α)-1/3 -1]

2

Three-

Zhuravlev,

dimensional

Lasokin,

diffusion

Tempelman equation 239 240

3. Results and Discussion

241

3.1 Thermogravimetric Analysis (TGA) 100

0

90 -0.005

70 60

-0.01

50 40

-0.015

30 20

Mass loss rate (%/°C)

80

Mass loss (%)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

-0.02

10 0

-0.025 50

150

250

350

450

550

650

750

850

Temperature (°C) TG/EFBF

TG/SLUDGE

DTG/EFBF

DTG/SLUDGE

242 243

Figure 1. TG and DTG curves of EFBF and POME sludge at heating rate of 20 °C/min

244 14 ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 37

245

Figure 1 shows the thermo-gravimetric data (TG) and thermo-gravimetric derivative (DTG)

246

curves of EFBF and POME sludge at a heating rate of 20 °C/min in nitrogen (N2)

247

atmosphere. As observed, the pyrolysis of EFBF could be distinguished into three stages: 1)

248

evaporation of moisture and light organic compounds (50 – 160 °C); 2) devolatilization of

249

mainly cellulose and hemicellulose (EFBF), or decomposition of heavy organic compounds

250

(sludge) (160 – 420 °C); and 3) decomposition of lignin and other stronger chemical bonds

251

(420 – 660 °C). The evaporation of moisture occurred below 160 °C, including the drying of

252

surface moisture to of free moisture, and then to of bound moisture

253

process could be deduced from the presence of a minor peak in the DTG curve in this stage at

254

81.51 °C. EFBF is a lignocellulosic biomass, consisting of cellulose (35 – 50 wt%),

255

hemicellulose (15 – 30 wt%), and lignin (12 – 35 wt%) 29. Within the temperatures of 160 °C

256

and 420 °C, the cellulose and hemicellulose devolatizes, contributing to the major peak

257

shown in the DTG curve at 347.16 °C. Towards the end of this stage of reaction, the sample

258

loses 63.57 % of the original mass, which is agreeable with the percentages of the cellulose

259

and hemicellulose present in EFBF. The results obtained were similar to that obtained by

260

Nyakuma et al. (2014), who reported that in this stage, the condensable and non-condensable

261

matters in the biomass were thermally decomposed into gases, char, and tar. Beyond 420 °C,

262

lignin decomposed at a comparatively slower rate 30. The slower decomposition rate of lignin

263

was because lignin is a complex natural polymer of aromatic compounds, which required a

264

higher temperature to degrade as compared to cellulose and hemicellulose. Similar to Idris et

265

al. (2010), no obvious weight loss was observed beyond 550 °C 13.

266

With regard to the decomposition of POME sludge, three stages were observed and was

267

similar to the results obtained by Thangalazhy-Gopakumar et al. (2015) 5. The first stage

268

mainly corresponds to the dewatering of the samples at a temperature below 210°C. The

269

second stage at 210 °C to 440 °C involves the main decomposition of the POME sludge.

28

. This dewatering


Page 17 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

270

Decomposition of major organic compounds present in sludge would be decomposed in this

271

temperature range. Beyond 440 °C, the decomposition rate of POME sludge reduced greatly

272

as carbonaceous materials formed from the pyrolysis of the sample in the second stage such

273

as tar and coke goes through secondary cracking. Apart from that, other inorganic materials

274

present in the POME sludge were being decomposed

275

the structure of inorganic compounds is more complex and their chemical bonds are more

276

difficult to break, contributing to the slower decomposition rate in stage three.

277

The pyrolysis of both EFBF and POME sludge were divided into three stages, with similar

278

temperature ranges. However, the higher rate of mass loss in EFBF shows a higher reactivity

279

than POME sludge. This is because EFBF has higher organic volatiles content than POME

280

sludge 8.

281

31

. As compared to organic compounds,

3.2 Synergic effect

282 283 284 285

Figure 2. TG and DTG curves of: (a) 10 % sludge, (b) 25 % sludge, (c) 50 % sludge, and (d) 75 % sludge. T indicates the theoretical values (dotted lines) whereas E indicates the experimental values (solid lines).


Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 37

286 287

Figure 2 depicts the TG and DTG curves of EFBF- POME sludge blends at a heating rate of

288

20 °C/min in the presence of N2, respectively. From the peak position and height of DTG

289

curves, the combustion reactivity of the samples can be studied, where the DTG peak height

290

is directly proportional to the reactivity, whereas the temperature in correspondence to the

291

peak height is inversely proportional to the reactivity

292

percentages in the blends, the profiles of both the TG and DTG curves shifted from EFBF to

293

sludge where combustion reactivity gradually decreased.

294

The understanding of synergic effect in the production of bio-oil via co-pyrolysis is important

295

as it is one of the main factors responsible in the measuring or determining improvements in

296

oil quality and quantity 7. Positive or negative synergic effect relies on the type and contact

297

between the components, duration of pyrolysis, temperature and heating rate, removal or

298

equilibrium of volatiles formed, and addition of solvents, catalysts, and hydrogen donors 7,33.

299

Since the effect is dependent on the type of feedstock as mentioned, the synergistic effects of

300

different co-pyrolysis feedstock would vary according to their composition and pyrolysis

301

behaviour. Operating conditions contribute to the varying of synergic effect during co-

302

pyrolysis as well.

303

In order to investigate the synergistic interaction between EFBF and POME sludge, the

304

theoretical and experimental values of blends during co-pyrolysis are presented in Figure 2 .

305

Comparing theoretical and experimental values, TG curves for blends showed differences in

306

terms of mass loss and could be clearly observed from the major peaks of DTG curves. The

307

experimental values showed higher mass loss as compared to those of theoretical values,

308

suggesting that more cellulose and hemicellulose were being degraded than expected. Metal

309

oxides in sludge ash have been proven to act as heterogeneous catalyst in the recent years 34,35

32

. With the increase of sludge


Page 19 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

310

. POME sludge in this study contains metal oxides such as CaO, ZnO, and MgO that might

311

have posed catalytic effect in the pyrolysis of EFBF.

312

A difference between experimental and theoretical values hinted the existence of interaction

313

between EFBF and POME sludge, known as the synergistic effect

314

obtained a positive synergic effect in the co-pyrolysis of legume straw and coal

315

study, the biomass was considered as a hydrogen donor, which aided the hydrogenation of

316

coal upon pyrolysis, resulting in some positive synergies. Apart from that, Fei et al. (2012)

317

mentioned that synergistic effect was greater when the contact between particles improves 38.

318

However, a close contact between particles may pose an inhibitive effect as they will fill in

319

the interspaces of each other

320

synergistic effect varies with the type of feedstock upon co-pyrolysis.

321

36

. Zhang et al. (2007) 37

. In that

36

. This contradicting theory further confirms the fact that

3.3 Kinetics of Co-pyrolysis by Vyazovkin Method

322

The kinetics of EFBF and POME sludge pyrolysis were determined to guide and optimize

323

production. Vyazovkin method is a non-linear method that uses integration technique. As

324

compared to other integral model-free methods, i.e. Ozawa, Vyazovkin method provides a

325

more accurate value of activation energy. This is because Vyazovkin method acknowledges

326

the strong variation of activation energy with the degree of conversion, whereas the other

327

methods do not

328

limited due to the limitation experienced by mass transfer at high conversions (above 80

329

wt.%) 24,40.

25,39

. Despite its accuracy, the application of Vyazovkin method remains


Energy & Fuels

300

250

Activation energy (kJ/mol)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 37

200

150

100

50

0 0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

Degree of conversion, α

330

EFB

10%

25%

50%

75%

POME sludge

331 332

Figure 3. Activation energies of EFBF, POME sludge, and sludge percentages in the blends with respect to degree of conversion

333

Figure 3 shows the activation energies of different samples with respect to the degree of

334

conversion. The activation energies for EFBF and other blends with different sludge

335

percentages showed significant increase from conversions () 0.1 to 0.3, and hereafter a

336

slight increase until equal to 0.7. On the other hand, the activation energy of sludge

337

increased with the increasing of conversion until reaching a maximum of 232.61 &v/!wx

338

when = 0.6. As seen from the TG curves in Figure 1 and Figure 2, hemicellulose and

339

cellulose devolatilized below 70% conversion, which corresponds to the average activation

340

energy of 195.25 ± 11.63 &v/!wx . Typical activation energies obtained for the

341

devolatilization of both hemicellulose and cellulose are 100 – 111 kJ/mol and 195 – 236

342

kJ/mol, respectively 41–43, which is comparable to the results obtained in this study.

343

From Vyazovkin analysis, there are not much variations in activation energies for conversion

344

0.4 to 0.7, which mainly accounts for liquid product in fast pyrolysis. To identify the kinetic

345

models involved for EFBF, sludge, and their blends, master-plots method was employed.


Page 21 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

346 347

3.4 Kinetic Models of Co-pyrolysis by Master-plots Method

348

Utilizing the pre-determined E values obtained along with the temperature measured as a

349

function of , experimental master plots of

350

°C/min, 10 °C/min, 20 °C/min, 30 °C/min, and 40 °C/min) were plotted. On the other hand,

351

theoretical master plots of

352

functions as presented in Table 3. At different heating rates, the experimental master plots

353

presented similar results, signifying that the kinetics degradation process of EFBF, POME

354

sludge, and the blends could be described by a single kinetic model

355

models assumed certain ideal physical and geometrical conditions, disagreement between the

356

idealized and real systems might occur

357

experimental master plots fitted third order kinetic model (F3) the best and is presented in

358

Figure 4. This finding suggested that the chemical reactions during degradation of the

359

samples are the rate limiting steps to this thermochemical conversion

360

percentages of POME sludge in the blends increase, the more the experimental master plots

361

deviated from the theoretical master plot. This variation may be contributed by the combined

362

effect of EFBF and POME sludge in decomposition mechanism.

()

(5.g)

h(9)

h(96.i )

against under various heating rates (5

against were plotted in accordance to various kinetic

44,45

. As these kinetic

46

. Among the different kinetic models, the

47

. However, as the


Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 37

363 ()

against and experimental master plots of

h(9)

364

Figure 4. Theoretical master plots of

365 366 367

against at various heating rates 5 °C/min, 10 °C/min, 20 °C/min, 30 °C/min, and 40 °C/min; (a) EFBF (b) 10 % sludge, (c) 25 % sludge, (d) 50 % sludge, (e) 75 % sludge, and (f) Sludge.

(5.g)

h(96.i )

368

369

For a depth analysis on the reaction and diffusion mechanisms during co-pyrolysis of EFBF

370

and POME sludge, the decomposition was divided into three stages based on DTG curve at a

371

heating rate of 20 °C/min.

372

3.5 Kinetics and Mechanisms of Co-pyrolysis by Coats-Redfern (CR) Method

373

Using Coats and Redfern (CR) method, the solid-state mechanisms involved in the pyrolysis

374

of EFBF, POME sludge, and their blends were determined. Based on the mechanisms that

375

gave the highest correlation coefficient (R2), the kinetic triplets of the stages were calculated 21 ACS Paragon Plus Environment

Page 23 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

376

and identified accordingly. A high R2 indicates that the kinetic model fits the data obtained

377

well. The kinetic triplets are activation energy (E), apparent pre-exponential factor (A), and

378

kinetic model. E is the minimum energy required to break chemical bonds between atoms,

379

and thus can be used to characterize the reactivity of the sample. A is a constant that is more

380

closely related with the structure of the material, whereas low factors would indicate a

381

surface reaction or a ‘’tight’’ complex; high factors would indicate a ‘’loose’’ complex

382

Based on TGA results, the samples were divided to three different stages and analysed

383

separately. The analysis results are then presented in Table 4. Stage 1 ranged from 50 to 200

384

°C; Stage 2 from 200 to 400 °C; and Stage 3 from 400 to 650 °C. The highest correlations

385

ranged between 0.9753 – 0.9992, indicating the reliability of the kinetic parameters obtained.

386

Figure 5 shows the plots of ln[()/ ; ] against 1/ that gave the highest correlation

387

coefficients for the samples.

388

Table 4. Kinetic parameters for EFBF, POME sludge, and their respective blends at a heating

389

rate of 20 °C/min Stages

Sample (EFBF

E (kJ/mol)

A, (1/s)

:

Kinetic

8,48

.

R2

model

POME sludge) 1

100 : 0

116.47

1.40×1018

F3

0.9910

90 : 10

96.49

7.21×1014

F3

0.9968

75 : 25

91.95

1.44×1014

F3

0.9948

50 : 50

80.96

1.02×1012

F3

0.9882

25 : 75

70.73

2.32×1010

F3

0.9795*


Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

0 : 100

2

100 : 0

90 : 10

75 : 25

50 : 50

25 : 75

0 : 100

Page 24 of 37

49.17

4.47×106

F2

0.9831

66.63

1.51×10b

F3

0.9467*

38.34

1.81×104

F3/2

0.9853

89.33

2.77×107

F1

0.9986

79.41

9.29×105

R3

0.9972

168.49

2.41×1013

D3

0.9976

81.20

4.25×10a

F1

0.9914

71.67

1.55×10g

R3

0.9768

194.16

7.63×1015

D5

0.9972

81.27

4.43×106

F1

0.9970

193.80

7.50×1015

D5

0.9971

71.92

1.68×10g

R3

0.9880

74.11

1.03×106

F1

0.9992

65.68

4.65×104

R3

0.9961

141.01

7.36×1010

D3

0.9968

68.97

3.25×105

F1

0.9990

61.00

1.62×104

R3

0.9940

131.69

9.77×109

D3

0.9952

63.47

8.39×104

F1

0.9967

27.05

1.94×10;

R3

0.9385#

154.20

1.53×1012

D5

0.9987


Page 25 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

3

100 : 0

167.91

1.55×1012

F3

0.9836

90 : 10

202.13

1.32×1014

F3

0.9785

75 : 25

176.80

2.64×1012

F3

0.9753

50 : 50

173.55

3.46×1012

F3

0.9774

25 : 75

169.51

1.25×1012

F3

0.9971

0 : 100

211.31

1.39×1015

F3

0.9790

390

* Even though the highest correlation values were not obtained for F3 model, this model was

391

identified as the dominant kinetic model. # Even though R3 model was not dominant for

392

sludge, this model was selected in order to compare with EFBF and other blends at stage 2

393

decomposition.


Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 37

394 395 396

Figure 5. Plots of ln[g(x)/T2] against 1/T that gave highest correlations for all samples (sludge% in the sample)

397 398

Stage 1 mainly involves the dewatering of samples, where the third order chemical reaction

399

model () = (1-α) was in dominion. The exceptions to this model is EFBF: POME sludge

400

of 25 : 75 and 0 : 100, which have higher correlation coefficients attributed to second order

401

() = (1-α) and one and a half order f(α) = 2(1-α)

402

orders reflect the different rates of reaction, where higher reaction order signifies higher rate

403

of reaction. A lower rate of reaction for the two exceptions might be caused by inorganic

H

3

;

2

3/2

respectively. The different reaction


Page 27 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

404

materials (ash content) present in POME sludge. At this stage, the rate determining step was

405

the chemical reaction.

406

Hemicellulose and cellulose devolatilizes at Stage 2, where the samples experienced a major

407

mass loss. At this stage, the solid-state reaction was not only controlled by chemical reaction,

408

but also the diffusion and phase boundary. The major mechanisms involved in Stage 2 were

409

first order kinetic () = (1 − ), power law in phase boundary reaction () = 3(1-α) ,

410

and diffusion mechanism either Jander equation

411

Zhuravlev-Lasokin-Tempelman equation () = (3/2)(1 − α)4/3 [(1 − α)1/3 − 1]

412

The rate determining mechanism for first order kinetic (F1) is the chemical reaction, where

413

there was an equal probability of nucleation at each active site

414

boundary reaction (R3) is a reaction controlled by movement of an interface at constant

415

velocity and at which nucleation occurs virtually immediately, so that the surface of each

416

particle is covered with a layer of the product 50. This function relates and t for a sphere

417

reacting from the surface inward and is usually assumed to be the governing conversion

418

model in the combustion of certain carbonaceous materials

419

and Zhuravlev, Lesokin, Tempelman equation (D5) are three-dimensional diffusion

420

mechanisms. Jander equation (D3) is for reactions in a sphere, where diffusion in all three

421

directions is equally important 49. In contrast, Zhuravlev-Lesokin-Tempelman equation (D5)

422

does not reveal the shape of the particle.

423

Some of the samples showed high correlation coefficients for kinetic models power law in

424

phase boundary reaction (R3) and Jander equation (D3), which involve reaction for spherical

425

symmetries, suggesting the presence of spherical particles in EFBF or POME sludge. Since

426

both of these kinetic models are attributed to EFBF but not POME sludge, it can be

427

concluded that the sphere particles were contributed by EFBF, and that the shape of POME

2/3

() = (3/2)(1-α)

2/3

[1-(1-α)

1/3 -1

]

or

−1

49

. Power law in phase

49,51

. Both Jander equation (D3)


Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 28 of 37

428

sludge particle is unknown. As like any heterogeneous reactions, three major mechanisms

429

that control pyrolysis were interphase reaction, diffusion and chemical reaction. However,

430

during co-pyrolysis, the presence of sludge was not able to confirm spherical shape for

431

particles. The solid-state mechanisms for decomposition of lignocellulosic materials have

432

also been studied by other researchers. Liu et al. (2002) discovered that decomposition of the

433

wood and leaf of fir plant obtained good linearity for first order reaction (F1), Jander equation

434

(D3), and power law in phase boundary reaction (R3) as well, which is similar to that

435

obtained for EFBF in this study 52. Three-dimensional diffusion mechanism is common in the

436

pyrolysis of lignocellulosic biomass. Guo and Lua (2000) found that for the pyrolysis of

437

extracted oil palm fibres at low temperature regimes, Jander equation (D3) is the effective

438

mechanism 53; Yorulmaz and Atimtay (2009) studied the effective mechanisms for untreated

439

pine samples and discovered that for two oxidation regions, the three-dimensional diffusion

440

mechanisms Valensi equation (D2) and Ginstling-Brounstein equation (D4) were in dominion

441

54

442

For stage 3, devolatilization of lignin occurs for all samples and the unanimous model

443

obtained was third order kinetic model (F3). Similar study on the pyrolysis kinetics of lignin

444

was carried out by Xie et al. (2013) and determined that pine cone lignin fitted third order

445

kinetic model (F3) best for Horowitz-Metzger method and second order kinetic model (F2)

446

for Coats-Redfern method

447

concluded that second order reaction mechanism (F2) fits well for raw alkali lignin pyrolysis

448

with R2 of 0.9970

449

0.9960.

450

According to Yorulmaz and Atimtay (2009), the fact that thermal analysis allows the fitting

451

of more than one kinetic model to the samples is non-favourable and was noticed in this study

452

as well

.

55

. Employing Coats-Redfern method, Bu et al. (2016), too,

56

. However, third order reaction mechanism also showed a good R2 of

54

. Thus, for further studies, combining TGA including dynamic and isothermal 27 ACS Paragon Plus Environment

Page 29 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

453

studies could be used to obtain the exact mechanisms and thermal constants of the

454

devolatilization process. In the current study, third order (F3) mechanism was assumed to be

455

the main mechanism responsible for the devolatilization of samples in both Stage 1 and Stage

456

3. As for Stage 2, first order reaction (F1) is seen as the main mechanism, accompanied by

457

power law in phase boundary reaction (R3) and Jander equation (D3) or Zhuravlev, Lasokin,

458

Tempelman equation (D5) mechanisms.

459

As observed from the Table 4, activation energies for the Stage 1 decreased with the increase

460

of POME sludge percentages. The increase in water content in the sample with the increase in

461

sludge percentages may have been the cause of this finding. Besides that, as the percentages

462

of POME sludge increases, the amount of volatiles decreases, which in turns lower the

463

activation energy of the sample.

464

As the percentages of POME sludge increased in the sample, the temperature at which stage

465

decomposition (Stage 2) starts increased. This signifies that the energy barrier that needs to

466

be overcome is proportional to the percentages of POME sludge in the sample. However, as

467

seen from Table 4, the E calculated decreases with the increase in POME sludge percentage,

468

which is contradictory. This finding had been obtained by Mu et al. (2016) as well as Du et

469

al. (2014) 8,57. It was explained that the percentage of volatiles in the samples played a role in

470

contributing to the activation energy, where with the increase in volatile percentage,

471

activation energy increases. As observed in Figure 1, both TG and DTG curves showed that

472

the devolatilization stage of EFBF is indeed much steeper than that of POME sludge,

473

indicating the much higher volatile percentage in EFBF. This explains the higher values of

474

activation energy for EFBF, despite the observed lower decomposition temperatures.

475

As compared to other stages, Stage 3 that mainly involved the devolatilization of lignin

476

showed the highest activation energy. The activation energy obtained was 183.53 ±


Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 30 of 37

477

18.46 &v/!wx . Decomposition of lignin required higher activation energy as compared to that

478

of other components as lignin has a more complicated structure and higher molecular weight.

479

Besides that, char that formed in Stage 2 might have contributed to the higher activation

480

energy due to secondary cracking.

481

Conclusions

482

There were three stages to the pyrolysis of both EFBF and POME sludge, namely dewatering,

483

devolatilization of cellulose and hemicellulose, and lignin decomposition stages, respectively

484

at similar temperature ranges. However, EFBF showed higher reactivity regarding the higher

485

mass loss rate as compared to that of POME sludge. Upon co-pyrolysis conducted from TGA,

486

a positive synergistic effect was observed in the experiments. Next, the kinetic models

487

involved in the pyrolysis of EFBF, POME sludge, and their blends were determined using

488

master-plots method. Third order kinetic model (F3) was identified as the most suitable

489

model. However, as the percentages of POME sludge in the blends increase, a deviation from

490

theoretical master plot was noticed. Upon dividing the degradation of the biomass into

491

distinct stages, diffusion and reaction mechanisms involved in co-pyrolysis of EFBF, POME

492

sludge, were further analyzed using Coats-Redfern method. For Stage 1, the dominant kinetic

493

model was a third order reaction; for Stage 2, a first order reaction with power law in phase

494

boundary reaction (R3), and diffusion mechanism either Jander equation (D3) or Zhuravlev,

495

Lasokin, Tempelman equation (D5); for Stage 3, a third order reaction (F3). Study on

496

diffusion mechanism revealed spherical shape for EFBF paticles, whereas POME sludge lack

497

on specific shape. As the percentages of POME sludge increased in the blends, activation

498

energy decreased, which implied a catalytic effect of sludge ash content in main degradation

499

stage.

500

Acknowledgement


Page 31 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

501

The authors would like to express sincere gratitude to Ministry of Higher Education for the

502

realization of this research project under the Grant FRGS/1/2015/TK02/UNIM/02/1.

503

However, only the authors are responsible for the opinion expressed in this paper and for any

504

remaining errors.

505 506

References

507

(1)

Energy Commission. National Energy Balance 2013. 2013, 98.

508

(2)

Chong, C.; Ni, W.; Ma, L.; Liu, P.; Li, Z. The use of energy in Malaysia: Tracing energy flows from primary source to end use. Energies 2015, 8 (4), 2828–2866.

509

510

(3)

Palm Oil Board 2013, 5–10.

511

512

(4)

Shuit, S. H.; Tan, K. T.; Lee, K. T.; Kamaruddin, A. H. Oil palm biomass as a sustainable energy source: A Malaysian case study. Energy 2009, 34 (9), 1225–1235.

513

514

Mpoc, L. T. O. N. Malaysian Palm Oil Council (MPOC) Official Website. Malaysian

(5)

Thangalazhy-Gopakumar, S.; Al-Nadheri, W. M. A.; Jegarajan, D.; Sahu, J. N.;

515

Mubarak, N. M.; Nizamuddin, S. Utilization of palm oil sludge through pyrolysis for

516

bio-oil and bio-char production. Bioresour. Technol. 2015, 178, 65–69.

517

(6)

bunches for fuel application. J. Phys. Sci. 2011, 22 (1), 1–24.

518

519

Abdullah, N.; Sulaiman, F.; Gerhauser, H. Characterisation of oil palm empty fruit

(7)

Abnisa, F.; Mohd, W.; Daud, A. W. A review on co-pyrolysis of biomass: An optional

520

technique to obtain a high-grade pyrolysis oil. Energy Convers. Manag. 2014, 87, 71–

521

85.

522 523

(8)

Mu, L.; Chen, J.; Yao, P.; Zhou, D.; Zhao, L.; Yin, H. Evaluation of co-pyrolysis petrochemical wastewater sludge with lignite in a thermogravimetric analyzer and a 30 ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

524

packed-bed reactor: Pyrolysis characteristics, kinetics, and products analysis.

525

Bioresour. Technol. 2016, 221, 147–156.

526

(9)

Shuang-quan, Z.; Xiao-ming, Y.; Zhi-yuan, Y.; Ting-ting, P.; Ming-jian, D.; Tian-yu,

527

S. Study of the co-pyrolysis behavior of sewage-sludge/rice-straw and the kinetics.

528

Procedia Earth Planet. Sci. 2009, 1 (1), 661–666.

529

(10)

Lin, Y.; Ma, X.; Yu, Z.; Cao, Y. Investigation on thermochemical behavior of co-

530

pyrolysis between oil-palm solid wastes and paper sludge. Bioresour. Technol. 2014,

531

166, 444–450.

532

(11)

Wang, X.; Deng, S.; Tan, H.; Adeosun, A.; Vujanović, M.; Yang, F.; Duić, N.

533

Synergetic effect of sewage sludge and biomass co-pyrolysis: A combined study in

534

thermogravimetric analyzer and a fixed bed reactor. Energy Convers. Manag. 2016,

535

118, 399–405.

536

(12)

Zhu, X.; Chen, Z.; Xiao, B.; Hu, Z.; Hu, M.; Liu, C.; Zhang, Q. Co-pyrolysis behaviors

537

and kinetics of sewage sludge and pine sawdust blends under non-isothermal

538

conditions. J. Therm. Anal. Calorim. 2015, 119 (3), 2269–2279.

539

(13)

Idris, S. S.; Rahman, N. A.; Ismail, K.; Alias, A. B.; Rashid, Z. A.; Aris, M. J.

540

Investigation on thermochemical behaviour of low rank Malaysian coal, oil palm

541

biomass and their blends during pyrolysis via thermogravimetric analysis (TGA).

542

Bioresour. Technol. 2010, 101 (12), 4584–4592.

543

(14)

Gil, M. V.; Casal, D.; Pevida, C.; Pis, J. J.; Rubiera, F. Thermal behaviour and kinetics

544

of coal/biomass blends during co-combustion. Bioresour. Technol. 2010, 101 (14),

545

5601–5608.

546

(15)

Page 32 of 37

Vyazovkin, S.; Burnham, A. K.; Criado, J. M.; Pérez-Maqueda, L. A.; Popescu, C.;


Page 33 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

547

Sbirrazzuoli, N. ICTAC Kinetics Committee recommendations for performing kinetic

548

computations on thermal analysis data. Thermochim. Acta 2011, 520 (1), 1–19.

549

(16)

1997, 49, 1493–1499.

550

551

(17)

(18)

Jiang, J.; Liu, Z.; Liu, Q. Synergetic Catalysis of Calcium Oxide and Iron in Hydrogasification of Char. Energy & Fuels 2017, 31 (1), 198–204.

554

555

Vyazovkin, S. Evaluation of activation energy of thermally stimulated solid-state reactions under arbitrary variation of temperature. February 1997, pp 393–402.

552

553

Vyazovkin, S. ADVANCED ISOCONVERSIONAL METHOD. Jounal Therm. Anal.

(19)

Lee, H.; Juan, J.; Yun Hin, T.-Y.; Ong, H. Environment-Friendly Heterogeneous

556

Alkaline-Based Mixed Metal Oxide Catalysts for Biodiesel Production. Energies 2016,

557

9 (8), 611.

558

(20)

Ninduangdee, P.; Kuprianov, V. I. Combustion of an oil palm residue with elevated

559

potassium content in a fluidized-bed combustor using alternative bed materials for

560

preventing bed agglomeration. Bioresour. Technol. 2015, 182, 272–281.

561

(21)

Sci. Tech 2011, 8 (1), 203–221.

562

563

Refaat, A. A. Biodiesel production using solid metal oxide catalysts. Int. J. Environ.

(22)

Mahadevan, R.; Adhikari, S.; Shakya, R.; Wang, K.; Dayton, D.; Lehrich, M.; Taylor,

564

S. E. Effect of Alkali and Alkaline Earth Metals on in-Situ Catalytic Fast Pyrolysis of

565

Lignocellulosic Biomass: A Microreactor Study. Energy & Fuels 2016, 30 (4), 3045–

566

3056.

567

(23)

and paper mill sludge. Appl. Energy 2010, 87 (11), 3526–3532.

568

569

Yanfen, L.; Xiaoqian, M. Thermogravimetric analysis of the co-combustion of coal

(24)

Pérez-Maqueda, L. A.; Criado, J. M. The Accuracy of Senum and Yang’s 32 ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

570

Approximations to the Arrhenius Integral. J. Therm. Anal. Calorim. 2000, 60, 909–

571

915.

572

(25)

Vyazovkin, S. Modification of the integral isoconversional method to account for variation in the activation energy. J. Comput. Chem. 2001, 22 (2), 178–183.

573

574

Page 34 of 37

(26)

Zhou, L.; Wang, Y.; Huang, Q.; Cai, J. Thermogravimetric characteristics and kinetic

575

of plastic and biomass blends co-pyrolysis. Fuel Process. Technol. 2006, 87 (11), 963–

576

969.

577

(27)

Vlaev, L.; Nedelchev, N.; Gyurova, K.; Zagorcheva, M. A comparative study of non-

578

isothermal kinetics of decomposition of calcium oxalate monohydrate. J. Anal. Appl.

579

Pyrolysis 2008, 81 (2), 253–262.

580

(28)

Bryś, A.; Bryś, J.; Ostrowska-Ligęza, E.; Kaleta, A.; Górnicki, K.; Głowacki, S.;

581

Koczoń, P. Wood biomass characterization by DSC or FT-IR spectroscopy. J. Therm.

582

Anal. Calorim. 2016, 126 (1), 27–35.

583

(29)

Kumar, R.; Tabatabaei, M.; Karimi, K.; Sárvári Horváth, I. Recent updates on

584

lignocellulosic biomass derived ethanol - A review. Biofuel Res. J. 2016, 3 (1), 347–

585

356.

586

(30)

Nyakuma, B. B.; Johari, A.; Ahmad, A.; Abdullah, T. A. T. Thermogravimetric

587

analysis of the fuel properties of empty fruit bunch briquettes. J. Teknol. (Sciences

588

Eng. 2014, 67 (3), 79–82.

589

(31)

model-free kinetics analysis. J. Hazard. Mater. 2009, 161 (2), 1208–1215.

590

591 592

Liu, J.; Jiang, X.; Zhou, L.; Han, X.; Cui, Z. Pyrolysis treatment of oil sludge and

(32)

Ghetti, P.; Ricca, L.; Angelini, L. Thermal analysis of biomass and corresponding pyrolysis products. Fuel 1996, 75 (5), 565–573. 33 ACS Paragon Plus Environment

Page 35 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

593

Energy & Fuels

(33)

sawdust in autoclaves. J. Anal. Appl. Pyrolysis 2013, 104, 341–352.

594

595

Johannes, I.; Tiikma, L.; Luik, H. Synergy in co-pyrolysis of oil shale and pine

(34)

Cheng, S.; Li, A.; Yoshikawa, K. High Quality Oil Recovery from Oil Sludge

596

Employing a Pyrolysis Process with Oil Sludge Ash Catalyst. Int. J. Waste Resour.

597

2015, 5 (2).

598

(35)

Nam, S.-B.; Park, Y.-S.; Yun, Y.-S.; Gu, J.-H.; Sung, H.-J.; Horio, M. Catalytic

599

application of metallic iron from the dyeing sludge ash for benzene steam reforming

600

reaction in tar emitted from biomass gasification. Korean J. Chem. Eng 2016, 33 (2),

601

465–472.

602

(36)

a free fall reactor. J. Therm. Anal. Calorim. 2014, 117 (2), 817–823.

603

604

(37)

(38)

Fei, J.; Zhang, J.; Wang, F.; Wang, J. Synergistic effects on co-pyrolysis of lignite and high-sulfur swelling coal. J. Anal. Appl. Pyrolysis 2012, 95, 61–67.

607

608

Zhang, L.; Xu, S.; Zhao, W.; Liu, S. Co-pyrolysis of biomass and coal in a free fall reactor. Fuel 2007, 86 (3), 353–359.

605

606

Quan, C.; Xu, S.; An, Y.; Liu, X. Co-pyrolysis of biomass and coal blend by TG and in

(39)

Criado, J. M.; Sánchez-Jiménez, P. E.; Pérez-Maqueda, L. A. Critical study of the

609

isoconversional methods of kinetic analysis. J. Therm. Anal. Calorim. 2008, 92 (1),

610

199–203.

611

(40)

Cortés, A. M.; Bridgwater, A. V. Kinetic study of the pyrolysis of miscanthus and its

612

acid hydrolysis residue by thermogravimetric analysis. Fuel Process. Technol. 2015,

613

138, 184–193.

614 615

(41)

Grønli, M. G.; Várhegyi, G.; Di Blasi, C. Thermogravimetric Analysis and Devolatilization Kinetics of Wood. Ind. Eng. Chem. Res. 2002, 41 (17), 4201–4208. 34 ACS Paragon Plus Environment

Energy & Fuels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

616

(42)

Várhegyi, G.; Antal, M. J.; Jakab, E.; Szabó, P. Kinetic modeling of biomass pyrolysis. J. Anal. Appl. Pyrolysis 1997, 42 (1), 73–87.

617

618

(43)

Varhegyi, G.; Antal, M. J.; Szekely, T.; Szabo, P. Kinetics of the thermal

619

decomposition of cellulose, hemicellulose, and sugarcane bagasse. Energy & Fuels

620

1989, 3 (3), 329–335.

621

(44)

Tanaka, H. Thermal analysis and kinetics of solid state reactions. Thermochim. Acta

1995, 267, 29–44.

622

623

(45)

Shuping, Z.; Yulong, W.; Mingde, Y.; Chun, L.; Junmao, T. Pyrolysis characteristics

624

and kinetics of the marine microalgae Dunaliella tertiolecta using thermogravimetric

625

analyzer. Bioresour. Technol. 2010, 101 (1), 359–365.

626

(46)

Sánchez-Jiménez, P. E.; Pérez-Maqueda, L. A.; Perejón, A.; Criado, J. M. Generalized

627

kinetic master plots for the thermal degradation of polymers following a random

628

scission mechanism. J. Phys. Chem. A 2010, 114 (30), 7868–7876.

629

(47)

Irmak Aslan, D.; Parthasarathy, P.; Goldfarb, J. L.; Ceylan, S. Pyrolysis reaction

630

models of waste tires: Application of Master-Plots method for energy conversion via

631

devolatilization. Waste Manag. 2017.

632

(48)

Sima-Ella, E.; Yuan, G.; Mays, T. A simple kinetic analysis to determine the intrinsic reactivity of coal chars. Fuel 2005, 84 (14), 1920–1925.

633

634

Page 36 of 37

(49)

Alshehri, S. .; Monshi, M. A. .; Abd El-Salam, N. .; Mahfouz, R. . Kinetics of the

635

thermal decomposition of γ-irradiated cobaltous acetate. Thermochim. Acta 2000, 363

636

(1), 61–70.

637 638

(50)

Mahfouz, R. M.; Al-Khamis, K. M.; Siddiqui, M. R. H.; Al-Hokbany, N. S.; Warad, I.; Al-Andis, N. M. Kinetic studies of isothermal decomposition of unirradiated and γ35 ACS Paragon Plus Environment

Page 37 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

639

irradiated gallium acetylacetonate: new route for synthesis of gallium oxide

640

nanoparticles. Prog. React. Kinet. Mech. 2012, 37 (3), 249–262.

641

(51)

López-Fonseca, R.; Landa, I.; Elizundia, U.; Gutiérrez-Ortiz, M. A.; González-

642

Velasco, J. R. A kinetic study of the combustion of porous synthetic soot. Chem. Eng.

643

J. 2007, 129 (1), 41–49.

644

(52)

Liu, N. A.; Fan, W.; Dobashi, R.; Huang, L. Kinetic modeling of thermal

645

decomposition of natural cellulosic materials in air atmosphere. J. Anal. Appl.

646

Pyrolysis 2002, 63 (2), 303–325.

647

(53)

Guo, J.; Lua, A. C. KINETIC STUDY ON PYROLYSIS OF EXTRACTED OIL

648

PALM FIBER Isothermal and non-isothermal conditions. J. Therm. Anal. Calorim.

649

2000, 59, 763–774.

650

(54)

Yorulmaz, S. Y.; Atimtay, A. T. Investigation of combustion kinetics of treated and

651

untreated waste wood samples with thermogravimetric analysis. Fuel Process.

652

Technol. 2009, 90 (7), 939–946.

653

(55)

Xie, H.; Yu, Q.; Duan, W.; Wang, K.; Li, X.; Shi, X. Pyrolysis characteristics and

654

kinetics of lignin derived from three agricultural wastes. J. Renew. Sustain. Energy

655

2013, 5 (6), 63119.

656

(56)

catalytic pyrolysis of lignin. RSC Adv. 2016, 6 (103), 100700–100707.

657

658

Bu, Q.; Lei, H.; Qian, M.; Yadavalli, G. A thermal behavior and kinetics study of the

(57)

Du, Y.; Jiang, X.; Lv, G.; Ma, X.; Jin, Y.; Wang, F.; Chi, Y.; Yan, J. Thermal behavior

659

and kinetics of bio-ferment residue/coal blends during co-pyrolysis. Energy Convers.

660

Manag. 2014, 88, 459–463.

661


Kinetics and Mechanisms for Copyrolysis of Palm Empty Fruit Bunch

Recommend Documents