carbonization-activation of sewage sludge for producing ...€¦ · 1.3.1 tar definition and...

Carbonization-Activation of Sewage Sludge for

Producing High Quality Gas and Sludge Char

Doctoral Dissertation

Young Nam Chun

Department of Environmental Science and Technology

Interdisciplinary Graduate School of Science and Engineering

TOKYO INSTITUTE OF TECHNOLOGY

ii

Doctoral Dissertation

Carbonization-Activation of Sewage Sludge for

Producing High Quality Gas and Sludge Char

Young Nam Chun

Department of Environmental Science and Technology

Interdisciplinary Graduate School of Science and Engineering

Tokyo Institute of Technology

Advisor: Professor Kunio Yoshikawa

i

Table of contents

Chapter 1 General introduction 1

1.1 Background 1

1.2 Pyrolysis and gasification 3

1.3 Tar Definition and problem 4

1.3.1 Tar definition and maturation mechanism 4

1.3.2 The tolerance of end-use devices for tar 6

1.4 Tar reduction technology 8

1.5 Object of the thesis 10

Chapter 2 Rotary type dryer for drying dewatered sludge 12

2.1 Literature review 12

2.2 Sludge drying process 13

2.3 Material and methods 15

2.3.1 Experimental apparatus 15

2.3.2 Experimental method 16

2.3.3 Data analysis 17

2.4 Results and discussion 18

2.4.1 Parametric screening studies 18

1) Effect in the rotating drum temperature 18

2) Effect of the sludge residence time 19

3) Effect of the dryer load 20

4) Emission of volatile compounds 21

2.4.2. Novel design for the rotary drum dryer 22

2.4.3. Mass and energy balance 24

2.5 Summary 28

Chapter 3 Pyrolysis and gasification performances of the dried sludge 30


ii


3.2.1 Experimental setup 31

3.2.2 Experimental procedure 32

3.2.3 Sampling and analysis method for products 34

1) Tar sampling and analysis 34

2) Sampling and analysis for producer gas 35

3) Sludge char analysis 35

3.2.4 Test setup and procedure for benzene adsorption 35


3.3.1 Effects of pyrolysis, gasification and carbonization-activation 36

1) Mass yield in product 38

2) Characteristics of producer gas 39

3) Characteristics of tar formation 41

4) Characteristics of sludge char 43

3.3.2 Verification of adsorptive tar removal from a continuous pyrolyzer 47

1) Test setup and procedure for the sludge char adsorption 47

2) Adsorption characteristics of biomass tar 48

3.4 Summary 50

Chapter 4 Designing and design verification of a plasma-catalyst reformer 52




4.2.2 Experimental methods 55



1) Effects of steam feed rate 58

2) Effects of catalyst bed temperature 59

3) Effects of total gas feed rate 60

4) Effects of input electric power 61

5) Effects of biogas content 61

iii

4.4 Summary 64

Chapter 5 Plasma reformer performance for tar destruction 66




5.2.2 Experimental methods 69


5.2.4 Reaction mechanism for tar destruction 72


5.3.1 Effects of light aromatic and PAH tars 73

1) Destruction for light aromatic tar 73

2) Destruction for light PAH tar 81

5.3.2 Verification of tar removal at a continuous pyrolyzer 87

1) Test setup for tar removal in biomass pyrolysis 87

2) Experimental results in the decomposition of biomass tar by the plasma reformer 88

5.3.3 Plasma reformer with an external oscillation 90

1) Test setup and procedure for tar destruction 90

2) Experimental results in the benzene tar decomposition by the EPOR 91

5.4 Summary 99

Chapter 6 Sequential carbonization-activation system including char production

and tar removal 101



6.2.1 Dried sludge for experiment 103

6.2.2 Carbonization-activation experiment 103


6.3.1 Combined carbonization-activator 107

6.3.2 Sequential in-line carbonization-activation system 112

iv

1) Characteristics of a combined carbonization-activator 112

2) Plasma reformer and adsorber characteristics 116

6.3.3 Process analysis for a sequential in-line carbonization-activation system 118

1) Mass and energy balance for a carbonization-activator 118

2) Mass and energy balance for a plasma reformer 123

3) Mass and energy balance for an adsorber 127

4) Performance analysis in view of the total energy balance for a sequential in-line

treatment system 131

6.4 Summary 135

Chapter 7 Conclusion 137

Reference 141

1

Chapter 1

General introduction

1.1 Background

It is widely accepted that interest in environmental issues is constantly increasing. At the same

time, environmental issues have gradually been broadened with concepts, such as sustainable

development, which implies not only ecological, but also economic and social responsibilities.

The handling of sewage sludge is one of the most significant challenges in waste water

management [1].

Sewage sludge is regarded as the residue produced by the waste water treatment process,

during which liquids and solids are being separated. Liquids are being discharged to aqueous

environment while solids are removed for further treatment and final disposal. The

constituents removed during the waste water treatment include grit, screenings and sludge [2].

Of the constituents removed by effluent treatment, sludge is by far the largest in volume,

therefore its handling methods and disposal techniques are a matter of great concern.

Sustainable sludge handling may be defined as a method that meets requirements of efficient

recycling of resources without supply of harmful substances to humans or the environment [3].

In the two last decades, waste water treatment has been a very important development

probably due to the increasing limitations in water disposal. Due to this increase, the amount

of sewage sludge has also increased in accordance with this development [4]. The amount of

sludge produced is affected in a limited scale by the treatment efficiency while the sludge

quality is strongly dependent on the original pollution load of the treated effluent and also on

the technical and design features of the waste water treatment process.

Recently, the waste water treatment process is utilizing the NPR (nitrogen and phosphorous

removal) as an advanced biological process for municipal waste water treatment that has

anaerobic, anoxic, and aerobic basins (Figure 1.1) [5]. The NPR process satisfies

simultaneous treatment of nitrogen and phosphorous being contrary to each other. The mixed

sludge of excess activated sludge and digested sludge will be used for this study after

dewatering by a centrifuge.

The general options, which are available for the sewage sludge treatment and disposal, are

agricultural use, land disposal and thermal treatment.

The agricultural use of raw sludge or other composting practices are the best way for using

this waste. However, significant amount of sewage sludge cannot be used as fertilizer due to

the high heavy metal content. For this type of sewage sludge, the land disposal is the only

possible application. Before disposal, sewage sludge has to be treated to eliminate the bacteria,

viruses and organic pollutants.

Thermal treatments (incineration, pyrolysis, gasification, etc.) are interesting techniques to

stabilize sewage sludge for disposal [6]. Thermal treatments sometimes have been classified

as a method of disposal but in fact, it is a method of stabilization because the final destination

of ashes generated is the landfill.

2

Figure 1.1 Process chart of a waste water treatment

The incineration of sewage sludge requires that the dried-solids content should be raised to

about 33% for the sludge to be autothermic. However the process is subject to more stringent

for air emissions, making flue-gas cleaning equipment which is a major item of capital

expenditure; incinerator ash can leach heavy metals unless vitrified, and this would further

increase the cost [7].

The pyrolysis has advantages over conventional incineration processes with respect to fuel

economy, energy recovery, and the control of heavy-metal emissions [8]. However, process

efficiency is affected by the sludge moisture content, such that co-pyrolysis with other wastes

has been recommended in order to increase the dry-solids content of the sludge.

And the gasification produces a single combustible gas which can only be readily locally,

whereas pyrolysis gives multiple products (some of which are liquid and can be transported

and used remotely). Capital and operating costs are similar, such that the principal difference

between the two processes lies in the product value [9]. The fact that gasification produces a

single clean product makes it more attractive than pyrolysis for installation at a sewage-

treatment works because the gas is easier to use than a combination of gas and oil.

On the other hand, carbonization and/or activation as the thermal treatment should be

proposed to convert biomass like sewage sludge into resources (sludge char) and energy

(producer gas). The carbonization generally means the pyrolysis process. The pyrolysis

generally produces gas and carbide (i.e., char) products, but the carbonization is the process to

make maximum carbide only through the pyrolysis process. And the activation includes

physical and chemical treatments to make micropores in the char which is produced by the

carbonization [10]. The physical treatment uses steam or CO2 gas as an activator.

In addition, the carbonization process additionally produces pyrolysis gas, and the steam

activation obtains hydrogen-rich gas by the reforming of the pyrolysis gas. However, tar

generated from the pyrolysis should be the matter to be treated. As tar condenses at low

temperature, it can cause clogging, pipeline corrosion, and aerosol formation during the post-

production phase. Furthermore, if it enters the engine, it can block inlet channels, cooler, and

filter element due to polymerization [11-13]. Consequentially, to use the producer gas as a

high quality gas produced from the carbonization-activation process, after treatment devices

should be needed for tar-to-gas conversion and tar adsorption removal.

Therefore, in this study, a combined carbonization-activation system is proposed for the

3

production of high quality sludge char and producer gas from sewage sludge and biogas

which are by-products in waste water plants. In addition, basic studies are conducted for

developing and verifying performances of the carbonization-activator and the plasma reformer.

1.2 Pyrolysis and gasification

Figure 1.2 represents the origin of the major products in both high pressure and low pressure

pyrolysis.

The solid products can be distinguished by their origins: charcoal retaining the morphology of

the original lignocellulosic; coke arising from continued thermolysis after the deposition of

liquids and organic vapors; soot from homogeneous nucleation of high temperature

decomposition products of hydrocarbons from the vapor phase.

The direct production of liquids is postulated to occur mainly at pressure above atmospheric.

At atmospheric pressure or bellow it is not clear whether a liquid phase exists, after breaking

of the main polymer covalent bonds, prior to volatilization of the main components of

biomass. Lignin is known to soften at rather low temperatures, and the charcoal, though

retaining structural features of the biomass, does shrink, which could indicate a plastic state

where pyrolysis products pass directly into a liquid state before devolatilization. The prompt

gases, produced from the direct formation of gaseous species by the primary pyrolysis

reaction, are primarily CO2, H2O, and CO. These are largely associated with the char forming

reactions.

The sequential transformation of the primary products in the vapor phase can be divided into

three stages. At the primary stage, the slight cracking reactions occur on the time scale of

Figure 1.2 Pyrolysis and gasification pathways [14]

4

about one second before substantial conversion to permanent gases occurs. The higher

molecular weight lignin products are cracked to oxygenates. The secondary stage is the

formation of secondary products characterized by CO, light olefins, and the formation of

aromatics, even from the carbohydrates. The regime is of interest, since high value olefins and

light aromatics are a desirable product slate. The third stage leads to the tertiary products

characterized by the polynuclear aromatics (PNAs). These products generally form only in

high temperature conversion processes such as gasification and combustion and generally in

low yield.

Gasification is a thermal conversion process in which solid fuel or biomass is converted into a

gaseous fuel. Contrary to combustion, gasification produces a gas that is combustible. The

gasification process results in a combustible gas, also called syngas or producer gas. This gas

can be used in many different ways for e.g. the production of heat, power, or liquid fuels. The

gas can also be used to replace natural gas. This back-end flexibility is one of the major

reasons for the popularity of gasification.

1.3 Tar Definition and problem

1.3.1 Tar definition and maturation mechanism

Tars are defined as a generic term comprising all organic compounds present in the producer

gas excluding gaseous hydrocarbons (C1~C5).

Different classifications for tars are found in literatures [15-17]. In general, these

classifications are based on: properties of the tar components, and the aim of the producer gas

application. The tar components can be segregated and classified into five classes based on

their chemical, condensation and solubility behaviors, as given in Table 1.1.

Table 1.1 Classification of tars [18]

Class Class name Tar components Representative compounds

1 GC undetectable

Tars

The heaviest tars;

Not detected by GC

None

2 Heterocyclic Tars containing hetero atoms;

Highly water-soluble compounds

Pyridine, phenol, cresols,

quinoline, soquinoline,

dibenzophenol

3 Light aromatic

hydrocarbons

(LAH)

Aromatic components; Light

hydrocarbons with single ring;

Important from the point view of tar

reaction pathways;

Not posing a problem on

condensability and solubility

Benzene, toluene,

ethylbenzene,

xylenes, styrene

4 Light polyaromatic

hydrocarbons

(LPAHs)

Two and three rings compounds;

Condensing at low temperature even

at very low concentration

Indene, naphthalene,

methylnaphthalene,

biphenyl, acenaphthalene,

fluorine, phenanthrene,

anthracene

5 Heavy polyaromatic

hydrocarbons

(HPAHs)

Larger than three-rings; Condensing

at high temperatures at low

concentrations

Fluoranthene, pyrene,

chrysene, perylene,

coronene

5

The presence of tars in the fuel gas is one of the main technical barriers in the biomass

gasification development. These tars can cause several problems, such as [19] cracking in the

pores of filters, forming coke and plugging the filters, and condensing in the cold spots and

plugging the lines, resulting in serious operational interruptions. Moreover, these tars are

dangerous because of their carcinogenic character, and they contain significant amounts of

energy which should be transferred to the fuel gas as H2, CO, CH4, etc. In addition, high

concentration of tars can damage or lead to unacceptable levels of maintenance for engines

and turbines. The tar levels and composition vary with pyrolyzer or gasifier type, process

conditions, and biomass type.

Generally, the classification of tar compounds is based on the work by Evans and Milne [14].

They used molecular beam mass spectrometry (MBMS) to identify different reaction regimes

during thermal processes like pyrolysis and gasification.

Three major product classes were identified as a result of thermal gas-phase tar conversion

reactions. (Refer Figure 1.3) The primary product, found in the reactor temperature range of

around 400℃, are characterized by the presence of oxygenated compounds. The secondary

products include phenolics and olefins, the formation temperature range here is 500~700℃.

Tertiary products appear in the reaction regime of more than 800℃ and are characterized by

aromatics. Sometimes, this class is further subdivided into the classes ‘alkyl tertiary products’,

like methylnaphthalene, toluene and indene, and the ‘condensed tertiary products’, which

include the polyaromatic hydrocarbons (PAH). A list of important single tar compounds and

their classification according to Evans and Milne is given by Milne et al. [15].

Figure 1.3 Tar maturation scheme proposed by Elliott [20]

The description of process changes should be seen as a function of the reaction severity,

which combines both temperature and time. Evans and Milne [14, 21] show the trade-off in

product distribution as a function of these two parameters by using multivariate analysis of

product composition. Another important factor is the importance of gas phase reactions

leading to tar synthesis. Hydrocarbon chemistry, based on free radical processes, occurs in

this thermal regime where olefins react to give aromatics. This process occurs at the same

time that dehydration and decarbonylation reactions cause the transformations shown in

Figure 1.3.

Evans and Milne [14, 21] used molecular beam mass spectrometry (MBMS) to suggest that a

systematic approach to classifying pyrolysis products as primary, secondary, and tertiary can

be used to compare products from the various reactors that are used for pyrolysis and

gasification. Four major product classes were identified as a result of gas phase thermal

cracking reactions:

① Primary products: characterized by cellulose-derived products such as levoglucosan,

hydroxyacetaldehyde, and furfurals; analogous hemicellulose-derived products; and

lignin-derived methoxyphenols;

6

② Secondary products: characterized by phenolics and olefins;

③ Alkyl tertiary products: include methyl derivatives of aromatics, such as methyl

acenaphthylene, methylnaphthalene, toluene, and indene;

④ Condensed tertiary products: show the PAH series without substituents: benzene,

naphthalene, acenaphthylene, anthracene/phenanthrene, pyrene.

The primary and tertiary products were mutually exclusive as shown by the distribution in

Figure 1.4. That is, the primary products are destroyed before the tertiary products appear. The

tertiary aromatics can be formed from cellulose and lignin, although higher molecular weight

aromatics were formed faster from the lignin-derived products [14, 21].

Figure 1.4 The distribution of the four tar component classes as a function of temperature [15]

The classification of the quantitatively analyzed tar compounds according to Milne et al. [15]

is shown in Table 1.2. It is used for all processes that convert the primary tar either into

secondary and/or tertiary tar.

1.3.2 The tolerance of end-use devices for tar [15]

For selecting an optimal integrated clean-up strategy, the intended end use (gas application)

for the pyrolysis and/or gasification gas is a key consideration. The most important end uses

can be summarized as follows:

Close-coupled combustion: kilns, ovens, furnaces, dryers, “town gas” for local

distribution, and boiler firing

Hydrogen fuel production

External combustion for power: externally fired turbines, Stirling engines, steam engines,

thermo-photovoltaic cells, catalytic oxidation, and thermo-electric systems

Internal combustion engines (ICE): Diesel and Otto engines

Compressors

Gas turbines

7

Fuel cells: molten carbonate, solid oxide, proton exchange membrane, and phosphoric

acid

Chemical synthesis: methanol, ammonia, methane, Fischer-Tropsch liquids, other

oxygenates.

Specifications for contaminant levels that can be tolerated in these end-use applications are

given in Table 1.3.

Table 1.2 Classification of quantitatively analyzed tar compounds [17]

Tar

compound

class

Compound type Compound name

Tar

compound

class

Compound type Compound name

Primary

tar

compound

Acids

Ketones

Phenols

Guaiacols

Furans

Acetic acid

Propionic acid

Butyric acid

Acetol (1-hydroxy-2-

proanone)

Phenol

2,3-Dimethylphenol

2,4/2,5-imethylphenola

2,6-Dimethylphenol

3,4-Dimethylphenol

3,5-Dimethylphenol

Guaiacol

4-Methylguaiacol

Furfural

Furfural alcohol

5-Methylfurfural

Secondary/

tertiary

tarb)

Monoaromatic

hydrocarbons

Miscellaneous

hydrocarbons

Methylderiva-

tives of

aromatics

Benzene

Ethylbenzen

a-Methylstyrene

3&2-Methylstyrene

4-Methylstryrene

3-Ethyltoluene

4-Ethyltoluene

2-Ethyltoluene

2,3-Benzofuran

Dibenzofuran

Biphenyl

Indene

2-Methylnaphthalene

1-Methylnaphtalene

Toluene

Secondary

tar

compound

Phenol

Monoaromatic

Hydrocarbons

Phenol

o-Cresol

p-Cresol

m-Cresol

p/m-Xylenea)

o-Xylene

Tertiary

tar

compounds

PAH:2-ring

3-ring

4-ring

5-ring

6-ring

Acenaphthylene

Acenaphthene

Fluorine

Naphthalene

Phenanthrene

Anthracene

Fluoranthene

Pyrene

Benz[a]anthracene

Chrysene

Benz[e]acephenanth-

rylene

Benzo[k]fluoranthene

Benzo[a]pyrene

Perylene

Dibenzo[ah]anth-

racene

Indeno[1,2,3-cd]py-

rene

Benzo[ghi]perylene

<Note> a)

Compounds lumped together for analysis; b)

There are several compounds that appear in the second

and in one of the other two classes as well. This demonstrates the evolutionary development and the somewhat

arbitrary boundaries for the three tar classes

8

Table 1.3 Contaminant constraints

Gas Application/End Use Tar Loading (mg/Nm3, ppm)

Close-Coupled Combustion Limits are large

Town-Gas for local

distribution

50~500 ppm

Externally Fired

Stirling Engines

Higher than for ICE;

Tolerating raw producer gas

Steam engines Similar to boilers

Internal Combustion Systems

SI and diesel

Max of 100 mg/Nm3

Direct-Fired, Industrial Gas

Turbines

Tolerance for condensing tars 0.05-0.5 ppm

Compressors 50-500 mg/Nm3

Fuel Cells MCFC-external reforming

C2H6-tolerant; C2H4-less than 0.25 vol.%; C2H2-less

than 0.2 vol.%; benzene-vol.0.5%; aromatics-0.5 vol. %

MCFC-internal reforming Total contaminants-less than 80 ppb

Solid-Oxide, internal reforming Carbon deposition is a problem unless air is added to the

biogas

1.4 Tar reduction technology Physical processes will continue to play a very important role for the successful commercial

implementation of pyrolysis and gasification. They constitute the basic arm for removing

most raw pyrolyzer and gasifier contaminants, including tar.

The tar is removed mainly through wet or wet-dry scrubbing. Coalescers, demisters, and cold

filtration are also necessary supplements. These are well-known commercial methods and are

easily designed and applied depending on the specific needs of any pyrolysis and gasification

processes. However, the main problem arising from tar scrubbing is that condensed tar

components are merely transferred into another phase (water or solids such as scrubbing lime),

which then has to be disposed of in an environmentally acceptable manner. The problems

associated with the management of these waste water or solid residues.

The raw gas leaves pyrolyzers or gasifiers at temperatures between 100℃ and 800℃. If hot

gas filtration and tar cracking and/or reforming conversion follows, the temperature should be

as high as possible. This is the case of physicochemical conversion of tar, which will be

covered in this study.

The use of dry, medium temperature, technologies for the physical removal of tar is not yet

envisaged. Fabric, ceramic, and metallic filters can remove near-dry condensing tar particles

from pyrolyzer or gasifier gas. They are based on the principle that liquid tar condensing at a

relatively high temperature will rapidly react to form solid species behaving as particulates

rather than tar. The reasons they have not been used are the following: They will be only

partially effective at temperature higher than 150℃; an important amount of tar will remain at

the gas phase and pass through the filter without being retained. If a near-liquid layer is

formed on the surface of the filtering material, its stickiness will cause considerable

9

mechanical problems and frequent failures. Both operating and capital costs seem to be very

high.

An alternative could be the use of relatively high temperature adsorption on activated carbon

granular bed filters. The method is proposed in Hasler et al. [22]. They mention that charcoal

or activated carbon are thermally stable up to 300℃. Since conventional fabric filters are

expected to exhibit a limited tar separation efficiency, an activated carbon filter can be

installed after a fabric filter unit to remove high boiling hydrocarbons and possibly phenols.

The filter is preferably made as a fixed bed with granular charcoal or activated carbon. The

temperature should be as low as possible (e.g. 120℃), but above the gas dew point. The tar

laden activated carbon can be recycled in the gasifier as an extra feedstock. However, no

information has been found in the literature for the tar adsorption characteristics of

carbonaceous adsorbents from biomass producer gas.

The biomass char was noticed to have a good catalytic activity for tar removal [19]. In a

downdraft gasifier, both fuel and gas flow downwards through the reactor enabling pyrolysis

gases to pass through a throated hot bed of char [23]. This results in cracking of most of the

tars into non-condensable gases and water [24]. The two stage gasifier developed by the

Technical University of Denmark (DTU) gives almost complete tar conversion (< 15 mg/Nm3)

[25]. The high tar removal is related to passing the volatiles through a partial oxidation zone

followed by a char bed.

The feasibility of the catalytic cleaning of producer gas from the biomass gasification is

mainly determined by economics [25]. The economics of the overall gasification process is

affected by the cost of the catalyst downstream of the biomass gasifier, lifetime of the catalyst,

and gas cleaning temperature. The attractiveness of biomass char for solving the tar problem

comes from its low cost, its catalytic activity for tar reduction and natural production inside

the gasifier. The last characteristic gives the biomass char the possibility to be integrated in

the gasification process itself. However, there are no significant data or comprehensive studies

that explain the performance of biomass char for tar reduction.

Plasma technology can be offered as an alternative solution for thermal and catalytic

treatment [26, 27]. This process is capable of very high destruction efficiency, similar to

incineration. Plasmas contain highly reactive species, such as electrons, ions and radicals

which can enhance chemical reactions. Generally, plasmas can be divided into two categories:

thermal plasmas and non-thermal plasmas. Non-thermal plasmas are low pressure plasmas

characterized by high electron temperatures, and low ion and neutral temperatures. Thermal

plasmas and non-thermal plasmas have been a subject of active researches for many years,

with the number of applications steadily growing. Investigations on reforming and destruction

of organic compounds are increasingly investigated [28-31]. For these two general types of

plasma discharges, it is impossible to simultaneously keep a high level of non-equilibrium,

high electron temperature and high electron density, while most plasma chemical applications

use high power and a high degree of non-equilibrium to support selective chemical processes.

However, these parameters can be achievable in gliding arc discharge which will be used in

this study. Gliding arc plasma has a complicated space-time structure including quasi-

equilibrium and non-equilibrium periods as well as a fast transition between them. The fast

transition is due to a specific ionization instability which has a strong non-linear behavior [32,

33]. It may be simply characterized by the presence of burning flames between two metal

10

electrodes. This gliding arc plasma is not complex, can be generated at atmospheric pressure,

and is able to produce energetic radical species, more active than non-thermal plasmas [33-

35]. It offers high energy efficiency and selectivity for chemical reactions. The excitation

energy can be delivered to specific molecules in the reacting gas mixture [36].

The gliding arc plasma technology has successfully been used to decompose various organic

compounds [37-39]. Pemen et al. [40] reported the use of gliding arc discharge to remove tar

downstream of a 1 kg/h biomass gasifier at varying temperatures. They found that tar

conversion increased with applied energy, but with tar conversion of at most 40%. This

technology appears to have great potential in removing tar from biomass gasification.

However, there have been relatively few works with regards to plasma cracking of tar.

1.5 Objective of the thesis

The general objective of the thesis is to propose a waste treatment process for the by-

products of sewage sludge and biogas from a wastewater treatment plant. To do this, the eco-

friendly system was suggested like the process (Figure 1.5) having two ways. First,

conversion treatment of the waste sewage sludge into clean fuel energy and high porosity

sludge char; Second, hydrogen-rich gas production from the biogas coming from the

anaerobic digesters.

For the conversion treatment of the waste sewage sludge, the sequential in-line

sludge treatment system was newly designed and verified the process, consisting of

the rotary drum dryer, the combined carbonization-activator, the plasma reformer and

the adsorber. In addition, development of each component including the total system

integration (refer following specific objective for each chapter) was conducted for

improving the process efficiency.

For the hydrogen-rich gas production for the biogas, the plasma-catalyst reformer

was designed and verified its performance.

The sludge char produced from the carbonization-activator was applied for an adsorbent in the

adsorber. The clean producer gas should be used for end-use devices (e.g., fuel cells, gas

engines, gas turbines, etc).

Figure 1.5 Flow chart of a sequential treatment system for waste sewage sludge

11

▌First specific objective is to develop a novel rotary drum dryer for drying dewatered sludge

from a centrifuge. The dried sludge was used as a feedstock to the combined carbonization-

activator.

In Chapter 2, the dryer was newly designed focusing on a rotary kiln body (the deflector, the

pickup flights, the internal screw vane and the cylinder core) and an inside rotating body (the

knife-like blades, the fork-like stirrers and the fan-like blades). The newly designed dryer

could improve the drying efficiency and the energy efficiency with lower volatile compounds,

compared to the conventional rotary dryers. For verifying the effectiveness of the sludge

drying, parametric screening studies were done by changing the rotating drum temperature,

the sludge residence time, and the dryer load.

▌Second specific objective is to investigate characteristics of pyrolysis, steam gasification

and carbonization-activation in a batch-type fixed bed reactor. Particular attention is given to

the characteristics of the solid and gaseous products including tar which may give damages in

downstream applications.

In Chapter 3, experiments were conducted with a bench scale wire-mesh reactor to know the

best method for producing high quality sludge char and producer gas, simultaneously.

Comparative analysis on the formation characteristics of products such as gas, tar, and char

were conducted for each case.

▌Third specific objective is to develop a novel gliding arc plasma, and verify its

performance for biogas reforming and tar destruction.

In Chapter 4, the designing and design verification of the plasma arc reformer was conducted

for reforming the biogas produced from digesters installed in sewage treatment plants. The

developed gliding arc plasma reformer was also used for the tar destruction described in

Chapter 5. In addition, a catalyst reactor was combined with the plasma reformer to improve

the reforming efficiency. A parametric screening study was conducted for the variables that

affect the biogas reforming of the plasma-catalyst reformer, and presented the optimal

operating conditions for hydrogen-rich gas production.

In Chapter 5, the developed plasma reformer was used for tar destruction. The plasma

reformer applied were a gliding arc discharge type which was designed in Chapter 4. For tar

destruction performance demonstration, benzene and anthracene were selected as a light

aromatic tar and a light PAH tar, respectively. Experiments were performed on the parameters

that affect the tar decomposition efficiency, and the optimal operation condition was presented.

▌Last specific objective is to suggest a sequential in-line treatment system for energy and

resource utilization of the dewatered sewage sludge, and to verify its performance for

producing clean producer gas and high quality sludge char.

In Chapter 6, an integrated system with in-line connection of a combined carbonization-

activator, a plasma reformer, and a fixed bed adsorber was developed. The combined

carbonization-activator produced sludge char, tar and producer gas. The plasma reformer was

set to improve producer gas yield by destructing tar released from the carbonization-activator.

The fixed bed adsorber filled with the sludge char was installed for adsorption of residual tar

from the plasma reformer.

12

Chapter 2

Rotary drum dryer for drying dewatered sludge In sewage sludge treatment, a drying process is important to reduce the weight of the sludge.

When used as a pretreatment, sludge drying contributes greatly to a reduction in running costs.

Thus, there is strong demand for an efficient sludge drying process.

In spite of a great number of industrial applications, their design remains empirical. Therefore,

a novel rotary drum dryer having an inside rotating body was designed and verified its

performance. Parametric screening studies were conducted in the rotating drum temperature,

sludge residence time and dryer load. And to illustrate the effectiveness of the novel design

for the rotary drum dryer, the experiment had been made in an optimum. The condition was

taken in having low energy input and almost 10% of moisture content.

The rotary drum dryer was used for sludge drying to be used in the test of the combined

carbonization-activator in Chapter 6.

2.1 Literature review

The quantity of treated sewage as well as the level of their treatment results in an increasing

amount of sewage sludge. Therefore, new solutions regarding sludge treatment, management,

and utilization are in demand [41, 42]. Sludge condensation and dewatering processes are no

longer sufficient to cope with the still growing amounts of sludge or to reach the required

standards.

The form of the product obtained after the dewatering process is barely acceptable by several

potential clients, including, among others, agriculture, forestry, as well as the power industry.

The product requires further transformation and more advanced treatment. This shall be the

task of the sludge drying process, understood as the thermal drying process in which thermal

energy is delivered to the sludge in order to evaporate water. The sludge drying process

reduces the mass and volume of the product, making its storage, transport, packaging, and

retail easier and also enables pyrolysis/gasification, co-combustion, and the incineration of

sludge.

Numerous researches [43-46] have been carried out at both bench and field scales to

investigate various parameters that determine the sludge drying for different types of dryers.

The paddle dryer [44, 47] was designed to provide a continuous process to produce an

environmentally friendly dry final product from sludge, e.g., sewage or fecal material. The

wedge-shaped configuration of the paddles and continuous opposite rotation of the twin shafts

make it possible for the paddles to squeeze the sludge continuously during the drying process.

Therefore, the sludge will not stick still to the paddle surface at the beginning of the drying

process.

The conveyer dryer [48] was designed to dry sewage sludge on the two-stage conveyer belt

using thermal wind drying of one mechanical power. Free water and interstitial water on the

dewatered sludge should be evaporated on the upper conveyor belt, and then the sludge drops

onto the lower conveyor belt to destroy the matters including surface water and bound water

(see Figure 2.1).

13

The screw dryer [49] has been developed for the thermal treatment of dehydrated, highly

viscous sewage sludge with moisture content exceeding 80 wt.%. The sewage sludge was

transported by the revolution of the cylinder conveyor together with the tumbling and mixing

action of the screw and lifters. The heating of the sludge is conducted efficiently by the

combination action of the conduction and convection modes together with the gas-agitation

process.

The multiple hearth dryer [50] consists of several hollow plates placed horizontally above

each other. The sludge is added to the dryer on the top-plate. A continuously rotating rake

mechanism transports the sludge from plate to plate. The dried sludge is evacuated at the

bottom. A small distance is kept between the rake and the plate in order to avoid abrasion and

contact between the rakes and plates. This results in a static layer of dried sludge on the plates.

The rotating disk dryer [51] consists of several rotating disks installed in a horizontal drum

having a heating jacket to supply heat into the drum by conduction heat transfer indirectly.

Incoming dewatered sludge from a sludge hopper is mixed with partial dried sludge to control

moisture content for protecting the operation in the glue phase state and a scrapper is installed

on the heat transfer surface to prevent the sticking of the sludge.

Although many types of dryers have been proposed and developed, each has several problems.

The paddle dryer, the conveyor dryer and the screw dryer cannot avoid the deposition of dried

sticky sludge on the surface of each dryer when operated for a long period. The multiple

hearth dryer has lower capacity in sludge treatment and has a limitation for reducing the

moisture content in dried sludge. The rotary disk dryer needs a long time to dry the sludge and

has difficulty in the maintenance due to its complicated structure. Also, the design for each

dryer remains empirical in spite of a great number of industrial applications.

In this study, therefore, a novel rotary drum dryer was developed to make up for the weak

points and to improve the drying efficiency for sewage sludge. Also, parametric studies were

carried out on the rotating drum temperature, the sludge residence time, the dryer load which

were known as important factors through our numerical simulation [52]. Through above tests,

the optimal operating conditions were determined and suggested. In addition, the effects of

the drying temperature on the emission rate of NH3, CO2, and O2 which are primary emission

components [53], were evaluated.

2.2 Sludge drying process

Sewage sludge is a high moisture content complex material composed of inorganic and

organic matter. The water in sewage sludge is classified into four categories: the free water,

the interstitial water, the surface water and the bound water.

The free water does not associate with solid particles and includes void water which is not

affected by the capillary force. The interstitial water is trapped crevices and interstitial spaces

of flocs and organisms. The surface water holds on the surface of solid particles by adsorption

and adhesion. The bound water chemically combines by hydration [54]. It is hard to

distinguish the free water and the bound water of sewage sludge clearly and the removable

moisture by mechanical dewatering is the free water and the interstitial water. The surface

water and the bound water are known as irremovable by mechanical dewatering [55].

14

As shown in Figure 2.2, the drying process of sewage sludge is divisible into the constant rate

period (A-B), the primary falling rate period (B-C) and the secondary falling rate period (C-D)

[56]. As compared to solids drying, sewage sludge have shorter constant rate period and

longer falling rate period, and the drying is processed with the primary critical moisture

content (point B) and the secondary critical moisture content (point C). Moreover, it is hard to

clearly distinguish the critical moisture content point between the constant rate period and the

falling rate period. Therefore, to increase the drying efficiency of sewage sludge, it is needed

to maintain the constant rate period long as the specific-surface area increases by changing the

cake state of sewage sludge into the particle state.

Figure 2.1 Combination of moisture content Figure 2.2 Drying curve for sewage sludge

Drying process of sewage sludge can be separated into three phase i.e. the pasty, the lumpy

and the granular phases as shown in Figure 2.3 [57].

When sludge is introduced into the dryer, its phase is pasty. With the decrease of the moisture

content as well as the volume of the sludge bulk, the sludge becomes the lumpy phase which

is more sticky and elastic. After this phase, the lumpy sludge bulk begins to break up, and the

granular phase begins. The granular phase is considered as a mono-dispersed particulate

phase.

Figure 2.3 Change of state through drying process

15

2.3 Material and methods

2.3.1 Experimental apparatus

Figure 2.4 represents the experimental setup for testing drying of dewatered sewage sludge.

The test rig consisted of a rotating drum dryer, a sludge feeding device, a hot gas generator, a

control panel, and a measurement and analysis line. The rotary drum dryer and the sludge

feeding device were newly designed for high drying efficiency and uniform feeding for sticky

dewatered sludge.

The rotary drum dryer consisted of a rotary kiln (rotating drum, internal screw vane and

pickup flights, deflector) and an inside rotating body (knife-like blades, fork-like stirers, fan-

like blades). The rotating drum (inner diameter is 267 mm, length is 1,315 mm) has an

insulator on its wall, and mounts a cylinder core having radial, longitudinally oriented vanes.

The internal screw vanes and pickup flights are attached inside the wall of the rotating drum,

and a deflector with a series of holes is installed near the hot gas inlet. A rotary shaft extends

axially through the center of the drum, and knife-like blades, fork-like stirers and fan-like

blades are mounted on the rotary shaft.

The sludge feeding device consisted of a sludge hopper, a sludge screw feeder and a paddle

plate feeder. The sludge hopper reserves the pasty dewatered sludge, and the screw feeder has

a screw to supply sludge constantly by controlling the revolution. Dewatered sludge is so

sticky that it is difficult to drop it down to the feed screw installed at the bottom of the hopper.

Therefore, the paddle plate feeder was designed to provide an effective feeding apparatus to

prevent tunneling [58] of the sticky sludge in the sludge hopper.

The hot gas generator consisted of a combustor and a gas burner to supply hot gas into the

rotating drum for direct heating.

The control panel controls the rotating drum, the paddle plate feeders and the screw feeder in

the sludge hopper, and the rotary shaft in the rotating drum.

Figure 2.4 Schematic diagram of the experimental setup

16

The measurement and analysis lines consisted of temperature measurement and gas analysis

equipment. Temperature was measured by K-type thermocouples (diameter 0.3 mm; Ni/Cr

10%) connected with a data analyzer (Fluke, Hydra Data Logger; 2625A). The thermocouples

were installed inside the rotating drum and the combustor, and at the hot gas inlet and the

drying gas outlet for continuous monitoring. A gas chromatography (Shimadzu; GC-14B) and

a UV/visible spectrophotometer (Biochrom, Ultrospec 2100 Pro) were used for measuring the

concentrations of CO2, O2 and NH3 in the by-product gas. The sampling line consisted of a

glass wool filter and a silica gel condenser to remove particle matter and water, respectively.

Particularly, NH3 was absorbed by boric acid (H3BO3) solution in an impinger bottle to be

measured by the UV/visible spectrophotometer.

2.3.2 Experimental method

For the developed dryer, parametric screening studies were conducted by changing the

rotating drum temperature, the sludge residence time, and the dryer load. And to illustrate the

effectiveness of the novel design for this rotary drum dryer, the experiment has been

conducted in the optimal condition which was taken through the parametric studies for the

best operating condition. The conditions are shown in Table 2.1.

The ranges for each parameter were taken as the conditions for which the moisture content in

the dried sludge was lower than 25%. And the optimal condition was fixed at having almost

10% moisture which is commonly used for pyrolysis or gasification [59, 60].

In addition, a test without the rotary shaft was conducted in other to show the effect of the

inside rotating body of stirring and blade members.

Table 2.1 Experimental conditions for each parameter

Variables Rotating drum

temperature (℃)

Sludge residence

time (min)

Dryer load

(kg/m3h)

Ranges for

parametric study 225~260 16~20 40~70

Optimal condition 225 17 55

For the precise experiment on sewage sludge drying, the screw feeder was calibrated. Figure

2.5 represents the calibration data measuring the supply amount of the dewatered sludge as a

function of the rotation speed of the screw feeder. As the revolution number increased, the

discharge amount of sludge increased linearly. The revolution numbers of 4~8 rpm in the

calibration chart correspond to the sludge feed rates of 49~86 g/min. Therefore, the range of

the dryer load was calculated to be 40~70 kg/m3h by Eq. 2.1.

The revolution of the rotating drum was fixed at 8~11 rpm, and that of the rotary shaft (i.e.,

inside rotating body) was fixed at 250 rpm.

The physicochemical characteristics of the dewatered sludge used in this study are shown in

Table 2.2. The dewatered sludge was analyzed for proximate and ultimate analyses. The

17

dewatered sludge was treated by a centrifuge to dehydrate raw sewage sludge. The moisture

content of the dewatered sludge was over 80%.

Figure 2.5 Calibration curve of the sludge feed rate

Table 2.2 Physicochemical characteristics of the dewatered sludge

Item

Proximate analysis Ultimate analysis

Moisture Volatile

matter Ash

Fixed

carbon C H O N S

Value

(wt.%) 81.3 66.8 25.5 7.67 46.9 6.8 33.6 8.4 4.3

2.3.3 Data analysis

The dryer load and the sludge residence time in the rotating drum were calculated as follows;

The dryer load was calculated by Eq. 2.1.

V

SD r

R

(2.1)

where DR (kg/m3h) is the dryer load, Sr (kg/h) is the sludge feed rate, and V (m

3) is the

volume of the rotating drum.

The sludge residence time was determined by Eq. 2.2.

RP

LRT

(2.2)

where RT (min) is the sludge residence time in the rotating drum, L (mm) is the length of the

rotating drum, P (mm/rev) is a pitch gap of the internal screw vanes per one rotation, and R

(rev/min) is the rotating speed of the rotating drum.

The dried sludge collected at the drying sludge outlet was checked for the moisture content,

and the drying efficiency was calculated;

The moisture content in the dried sludge was calculated by Eq. 2.3.

100W

WWMC

1

21

(2.3)

Revolution number (rpm)

Slu

dg

efe

ed

rate

(g/m

in)

2 4 6 8 100

30

60

90

120

150

y=10.397x+6.4918

R2=0.9921

18

where MC is the moisture content (%), W1 is the weight of dried sludge from the dryer (g),

and W2 is the weight after drying in an electric oven (maintaining at 105~110℃) for 4 h (g).

The drying efficiency was calculated by Eq. 2.4.

100WS

WC-WSDE (2.4)

where DE (%) is the drying efficiency, WS (%) is the moisture content of the dewatered

sludge and WC (%) is the moisture content of the dried sludge.

2.4 Results and discussion

For keeping a set moisture content in the dried sludge, maintaining a constant temperature in

the rotary drum dryer is important. To show the stabilization of the temperatures at each part

of the experiment, Figure 2.6 depicts the temperatures at the optimum test condition (as

shown in Table 2.1).

The rotating drum was heated up to 340℃. After feeding the dewatered sludge at the

dewatered sludge feed point, the temperature was decreased and then maintained uniformly at

255℃. After this steady state, the dried sludge and the by-product gas were sampled from a

starting test point for analysis.

Figure 2.6 Stabilization procedure at the optimal condition

The results of the parametric test presenting the effects of the rotating drum temperature, the

sludge residence time, and the dryer load are described in section 2.4.1. The effectiveness of

the novel design was illustrated with the results of the optimal conditions in section 2.4.2. In

addition, the mass and heat balances were shown in section 2.4.3.

2.4.1 Parametric screening studies

1) Effect of the rotating drum temperature

Elapsed time (min)

Te

mp

era

ture

(Co)

0 50 100 150 200 250 300 350 4000

50

100

150

200

250

300

350

400

450

500

550

600

650

700

Starting test pointfor samplings

1 Combustor

Dewatered sludgefeed point

3 Rotating drum

4 Drying gas outlet

2 Hot gas inlet

19

Figure 2.7 represents the results according to the variation of the rotating drum temperature in

the range of 225~260℃. The sludge residence time and the dryer load were fixed at 17 min

and 55 kg/m3h, respectively.

As the rotating drum temperature increased, the moisture contents in the dried sludge were

reduced from 24% at 225℃ to 6.4% at 260℃. The reason for this is that the water in the

dewatered sludge gradually evaporated. With only a difference of 35℃ in the drum

temperature, a difference of 17.6% in the moisture content was exhibited. This means that the

temperature should be a key factor to control the moisture content in sludge drying.

Meanwhile, the drying efficiency increased from 69.7 to 91.5% due to decreasing water

content while increasing the inside temperature of the rotating drum.

The dried sludge derived by the dryer could be used for pyrolysis or gasification to produce

synthesis gas, oil and char [61, 62].

Particularly for the case of allowing the high-porosity char to be used as an adsorbent material,

pre-drying for the raw sludge should be an important factor. This is because the drying in the

dryer gave the preliminary development of a micropore structure [63]. Therefore, higher

drying efficiency can give a positive effect on higher porosity for the pre-dried sludge char.

Figure 2.7 Effect of the rotating drum temperature

2) Effect of the sludge residence time

Figure 2.8 presents the results after changing the sludge residence time from 16 to 20 min

with the fixed conditions at 255℃ of the rotating drum temperature and 55 kg/m3h of the

dryer load.

As the sludge residence time in the rotating drum increased, the moisture content at the

residence time of 16 min was 12.4%. At the residence time of 20 min, the level of the

moisture content decreased to 1.4%.

The drying efficiency gradually increased from 84 to 98.2% while the residence time

increased from 16 to 20 min. This could be explained that an increase in the residence time

causes the increase of contact time between hot gas and sludge bulk in the rotating drum.

Rotating drum temperature (oC)

Mo

istu

reco

nte

nt

(%)

Dry

ing

eff

icie

ncy

(%)

220 225 230 235 240 245 250 255 260 2650

5

10

15

20

25

30

50

60

70

80

90

100

Drying efficiency

Moisture content

20

Figure 2.8 Effect of the sludge residence time

3) Effect of the dryer load

Figure 2.9 shows the results according to the variation of the dryer load in the range of 45~85

kg/m3

h with 17 min of the sludge residence time and 255℃ of the rotating drum temperature.

The drying efficiency gradually decreased from 91.5 to 81.6% when the dryer load increased

from 45 to 85 kg/m3

h.

The moisture content at the dryer load of 45 kg/m3h was 6.8%. Higher moisture content,

14.7%, was shown at the dryer load of 85 kg/m3h. The increase of the moisture content when

increasing the dryer load is due to the overload of sludge compared to the capacity of the

dryer. Therefore, the mixing between hot gas and the drying sludge bulk was difficult, such

that the drying sludge could not be dried effectively.

Figure 2.9 Effect of the variation of dryer load

Sludge residence time (min)

Mo

istu

reco

nte

nt

(%)

Dry

ing

eff

icie

ncy

(%)

16 17 18 19 200

3

6

9

12

15

70

75

80

85

90

95

100

Drying efficiency

Moisture content

Dryer load (kg/m3h)

Mo

istu

reco

nte

nt

(%)

Dry

ing

eff

icie

ncy

(%)

45 55 65 75 850

5

10

15

20

80

85

90

95

100

Drying efficiency

Moisture content

21

4) Emission of volatile compounds

The volatile compounds (VCs) produced during the sewage sludge drying process were

measured at the drying gas outlet, and the results are shown in Figure 2.10. The experiments

were achieved by varying the rotating drum temperatures from 225 to 260℃.

It was found that CO2 and NH3 were the primary components released from the sewage

sludge drying process, similar to the result of Deng et al. [53].

The dryer was directly heated by hot gas which was produced from a hot gas generator.

Propane was used as a fuel with the fixed air ratio of 1.4. The mixture was emitted from a

commercial Bunsen burner, generating the combustion gas of CO2 and O2 shown as “base line”

in Figure 2.10.

The CO2 gas measured during the drying process included both the volatile gas emitted from

sludge drying and combustion gas from fuel burning. The total CO2 concentration increased

from 8.9 to 12.1% when the rotating drum temperature increased, while CO2 in the

combustion gas maintained almost constant value of 8.4% due to the fixed air ratio. Therefore,

it should be known that the volatile CO2 increased from 0.6 to 3.6%. The volatile CO2 emitted

from the sludge drying was formed from the hydrolysis of amino acid, and the other method

is that it was formed from the decomposition of bicarbonate [64].

The dehydration process was carried out at a low temperature (T<200℃), where the free

water and some light organic compounds such as CO2 and CH4 were released. In the region

200℃<T<350℃ , the major decomposition or depolymerization occurred, which was

accompanied by the decrease of carbon and water and the increase of CO2 and CH4 in the

products [65]. The released CH4 may oxidize to CO2 and H2O with the consumption of O2.

That is why O2 slightly decreased with increasing the dryer temperature.

NH3 increased from 0.24 to 0.65 ppm when increasing the rotating drum temperature. The

NH3 emitted from sludge drying is formed through the hydrolysis of protein. When the

protein in sludge dissolves, it hydrolyzes to form multipeptide, dipeptide, and amino acid. The

amino acid further hydrolyzes to form organic acid, NH3 and CO2.

Figure 2.10 Emission of volatile compounds

22

Through the parametric study, the highest efficiency in the rotary drum dryer was shown

under the conditions in which the rotating drum temperature and sludge residence time were

increased and the dryer load was decreased.

However, it should be considered wholly in view of a high drying efficiency, a high energy

efficiency, and lower volatile compounds (particularly light organic compounds such as CH4)

to be used in thermal treatments such as pyrolysis and gasification. To satisfy these conditions,

the moisture content should be close to 10% [61, 62]. Therefore, the best operating conditions

for the developed dryer were set at 255℃ of the rotating drum temperature, 17 min of the

sludge residence time, and 55 kg/m3h of the dryer load. The average diameter of the dried

sludge created about 8mm and weight reduction was 80%. The drying efficiency and the

moisture content in the dried sludge were 84.8 and 12.4%, respectively.

2.4.2. Novel design for the rotary drum dryer

To verify the effects of the novel design, an experiment at the optimal conditions (refer to

Table 2.1) was conducted. The results are illustrated in Figures 2.11 and 2.12.

Figure 2.11 shows the state change before and after sludge drying by the novel drier.

Compared to dewatered sludge (a) which was treated by the centrifuge, the dried sludge (b)

features smaller size and volume. The average diameter of the dried sludge created was about

8 mm, and the weight reduction was 80%.

The drying efficiency and the moisture content in the dried sludge were 84.8 and 12.4%,

respectively.

(a) Dewatered sludge (b) Dried sludge

Figure 2.11 Comparison of the dewatered sludge and dried sludge

Figure 2.12 shows the dried sludge treated by the rotary drum dryer without an inside rotating

body, and comparisons of the moisture content and average sludge diameter for the dryers

with and without an inside rotating body.

In case of the sludge dried without an inside rotating body, the drying sludge in the rotating

drum has been gradually agglomerated. Therefore, the average diameter and the moisture

content were increased to 18mm and 26.7% respectively, compared to 8mm and 12.4% with

the inside rotating body. It could be known that the inside rotating body, which was designed

23

by this study, is very effective to crush the agglomerated sludge and to evaporate sludge water

by convective heat transfer for the optimum design.

(a) Without inside rotating body (b) Comparisons of moisture content

and average diameter

Figure 2.12 Comparison of dryers with and without the inside rotating body

As already explained, the rotary drum dryer (refer to Figure 2.4) was a new design,

particularly in the rotary kiln body (deflector, pickup flights, internal screw vane, cylinder

core) and inside rotating body (knife-like blades, fork-like stirrers, fan-like blades) for

improving drying efficiency.

The deflector located adjacent to the hot gas inlet deflects the flow direction of the hot drying

gas so that the drying gas may flow evenly within the rotating drum to heat the dewatered

sludge and, thereby, to suppress the sludge deposition on the inner wall.

The pickup flights provide an inner surface of the drum for carrying the sludge upwardly

incident to the rotation of the drum. At the same time, as the drum rotated, the internal screw

vane conveyed in-process sludge from the sludge inlet to the dried sludge outlet. With such a

construction, it is possible to mix efficiently in-process sludge and drying gas from the hot gas

inlet while moving to the dried sludge outlet. However, the sludge was dried as aggregated for

a relatively long period of time, so that the aggregate mass of sludge was discharged from the

dryer with only the surface dried and with the interior thereof incompletely dried, as shown by

the product in Figure 2.12(a).

Therefore, stirring members were mounted on the inside of the rotating body in order to break

the sludge charged into the drum in the proximity of the sludge inlet. The knife-like blades

can slice the pasty (phase) sludge coming from the sludge inlet through the screw feeder. And

the fork-like stirrers prevents the sticking of the sticky sludge (lumpy phase) to the drum wall

and inside of the rotating body by stirring the sticky sludge. And, further, the drying gas is

caused to swirl by the stirring members to form turbulent flow so that sludge may be dried

efficiently, even when the sludge is hard to dry. The fan-like blades adapt to the lower

velocity of the drying gas flowing within the rotating drum toward the drying gas outlet

incident to the rotation of the rotary shaft. Therefore, the sludge particles are prevented from

being discharged to the outside of the drum while being entrained in the drying gas. And,

thereby, the concentration of the sludge particles residing within the drum is increased or the

average residence time of the particles in the drum is prolonged in order to enhance the drying

efficiency.

24

The cylinder core modifies particle paths by excluding the flow from the elongated, generally

longitudinally oriented, portion of the interior of the drum and prevents short, circulated

passage of the drying gas directly through the drum. In addition, the vanes rotating with the

cylinder core divert at least some in-process sludge away from and along the core and impart

a rotational or spiral aspect to their path. This should be the best conditions for the sludge to

be dried effectively and crushed finely as the granular (phase) sludge at the final drying stage.

In the end, a good product could be taken as being like the dried sludge of Figure 2.11(b).

Moreover, the rotary drum dryer could achieve high drying efficiency with low energy

consumption and low particle emission.

In addition, to verify the effectiveness of this novel rotary drum dryer, it was compared with a

steam dryer developed by Mtsuo et al. [43]. As can be seen in Figure 2.13, the drying rate of

the novel dryer was higher than that of the steam dryer. This is because the novel dryer was

effectively designed for the task of sludge drying.

Figure 2.13 Comparison between this novel dryer and a steam dryer developed

by Mtsuo et al. [43]

2.4.3. Mass and energy balance

The mass and energy balances were calculated to clarify the energy requirement and energy

loss in the rotary drum dryer.

Figure 2.14 represents the mass balance in the combustor, the rotary drum dryer and the dust

collector. The combustor produces hot gas by burning fuel (C3H8) at Bunsen burner. The dry

hot gas (14.16 kg/h) comes into the rotating drum, and directly contacts with feeding bone dry

sludge (0.816 kg/h). The dried solids (0.784 kg/h) drops down to a drying sludge outlet, and

the dry exhaust gas (19.99 kg/h) goes out through a dust collector which collects dry particles

(0.032 kg/h) at the dust collector.

Dry

ing

rate

(g-H

2O

/kg

-DS

s)

0

0.2

0.4

0.6

0.8

1

Steam dryerNovel dryer

0.513

0.645

25

Figure 2.14 Diagram of mass balance

The heat balance was calculated, and the result is shown in Table 2.3.

▌Heat input was calculated by Eq. 2.5.

)T(TCGH refinpcwin (2.5)

where Hin is the heat input to a rotary drum dryer (kJ/h), Gcw is the amount of combustion gas

(kg/h), Cp is the specific heat of combustion gas (kJ/kg℃), Tin is the temperature of hot gas

(℃), and Tref is the temperature of ambient air (℃).

▌Heat output was calculated by Eqs. 2.6 to 2.13 as heat for liquid evaporation, heat for

outgoing solid product, heat loss in radiation, and heat loss in exhaust.

▰ Heat for liquid evaporation [66]

(i) Heat for vaporization of liquid (Hlv; kJ/h)

edustindusteoutinoutlv L)m(mWL)m(mWH (2.6)

where Wout is the bone dry sludge output from a dryer (kg/h), min is the moisture in feed

sludge (kg/kg of bone dry sludge), mout is the moisture in dried sludge output (kg/kg of dry

sludge output), Le is the latent heat at reference temperature (kJ/kg), Wdust is the dry solid

output from a dust collector (kg/h) and mdust is the moisture in dried particle output (kg/kg of dry

solid output).

(ii) Heat for superheating of evaporated vapor up to exhaust gas temperature (Hesup; kJ/h)

)T(TC)m(mW

)T(TC)m(mWH

refoutpvdustindust

refoutpvoutinoutesup

(2.7)

where Cpv is the specific heat of vapour (kJ/kg℃) and Tout is the temperature at the exit from

26

the rotary drum dryer (℃)

▰ Heat for outgoing solid product

(i) Heat given to drying sludge (Hs; kJ/h) [67]

)T(TCW)T(TCWH in-sdust-spsdustin-sout-spsouts (2.8)

where Cps is the specific heat of dried sludge (kJ/kg℃), Ts-out is the temperature of dried

sludge (℃), Ts-in is the temperature of dewatered sludge (℃), and Ts-dust is the temperature of

sludge particles in the dust collector (℃)

(ii) Heat for moisture in dried sludge (Hmd; kJ/h)

)T(TCmW)T(TCmWH refdust-sp-pldustdustrefout-sd-ploutoutmd (2.9)

where Cpl-d is the specific heat of water at the temperature of drying sludge outlet (kJ/kg℃),

Cpl-p is the specific heat of water at the temperature of the dust collector (kJ/kg℃)

▰ Heat loss in radiation (HR; kJ/h)

The heat loss in radiation can be calculated by getting the difference between the absorbed

heat from hot gas flow (Hg) and the utilized total heat (Htotal).

totalgR HHH (2.10)

Heat absorbed by the dryer from hot gas flow before exhaust (Hg; kJ/h)

)T(TCGH outinpcwg (2.11)

Total heat utilized in the dryer (Htotal; kJ/h)

mdsesuplvtotal HHHHH (2.12)

▰ Heat loss in exhaust (Hexh; kJ/h)

)H(HHH Rtotalinexh (2.13)

▌Thermal efficiency can be calculated by Eq. 2.14, and its value was 73.8%.

100H

HH(%) TE

in

esuplv

(2.14)

▌Typical specific energy consumption can be calculated by Eqs. 2.15 and 2.16, and its value

was 3.49 MJ/kg of water.

frin W / H water)of (MJ/kg SEC (2.15)

where the water feed rate (Wfr) can be calculated from Eq. 2.16.

ininfr Wmkg/h)(W (2.16)

Figure 2.15 represents an energy flow diagram. The heat input energy supplied to liquid and

solids is 76% of the total heat input. So, the energy used for evaporating the liquid in the

sludge by heating is 73.8%, and the energy used for heating the sludge solids including

moisture is 2.2%. Others are the heat losses by radiation (11.9%) and exhaust gases (12.1%).

27

Table 2.3 Energy balance in the rotary drum dryer

Parameter Unit Quantity Percent of

heat input Symbol

Heat input to a rotary drum dryer kJ/h 11,320 100 Hin

Heat out from a rotary drum dryer

( = I + II + III + IV ) kJ/h 11,320 100 Hout

I. Heat for liquid evaporation

(i) Heat for vaporization of liquid

(ii) Heat for superheating of

evaporated vapor up to exhaust

gas temperature

kJ/h

kJ/h

7,811

540.1

69

4.8

Hlv

Hesvp

Sub Total I kJ/h 8,351.1 73.8 Thermal efficiency, TE

II. Heat for outgoing solid product

(i) Heat given to drying sludge

(ii) Heat for moisture in drying

sludge

kJ/h

kJ/h

234.16

22.61

2.0

0.2

Hs

Hmd

Sub Total II kJ/h 256.77 2.2 Heat loss in outgoing

material

III. Heat loss in radiation;

Hg – (Subtotal I + Subtotal II) kJ/h 1,342.13 11.9 Radiation loss, HR

IV. Heat loss in exhaust;

Hin – (Subtotal I + Subtotal II + HR) kJ/h 1,369 12.1 Exhaust loss, Hexh

Figure 2.15 Energy flow diagram

28

Comparison of the novel rotary drum dryer to conventional other dryers are summarized in

Table 2.4. The novel rotary drum dryer developed in this study had highest typical drying

efficiency as 84.4%, while the specific energy consumption calculated was 3.49 MJ/kg of

water which is mostly the lowest value.

Table 2.4 Comparison of the novel rotary drum dryer to conventional other dryers [66]

Dryer group and type Typical heat loss

sources

Typical specific energy

consumption

( MJ/kg of water )

Drying

efficiency

(%)

Novel rotary drum dryer

(at optimal operating condition)

Surface,

exhausts, leaks 3.49 84.8

Rotary

Indirect Rotary Surface 3.0 to 8.0 28 – 75

Cascading Rotary Exhausts, leaks 3.5 to 12.0 19 – 64

Band, Tray & Tunnel

Cross circulated tray/oven/ band Exhaust, surface 8.0 to 16.0 14 – 28

Cross circulated shelf /tunnel Exhaust, surface 6.0 to 16.0 14 – 38

Through circulated tray /band Exhaust 5.0 to 12.0 19 – 45

Vacuum tray/ band / plate Surface 3.5 to 8.0 28 – 64

Drum Surface 3.0 to 12.0 19 – 75

Fluidized /Sprouted bed Exhaust 3.5 to 8.5 28 – 64

Spray

Pneumatic conveying/Spray Exhaust 3.5 to 8.0 28 – 64

Two stage Exhaust, surface 3.3 to 6.0 38 – 68

Cylinder Surface 3.5 to 10.0 23 – 64

Stenter Exhaust 5.0 to 12.0 19 – 45

The main target of this chapter is to develop the novel rotary drum dryer so that the dried

sludge will be used for the tests of the combined carbonization-activator in Chapter 6. So,

according to the target, the development was interested and concentrated only to the dryer

itself, not overall system. So, for the our tests, the odor in the VCs was temporarily removed

by using commercial burner at the exit of the dryer.

However, for the overall system in the sludge drying, the after treatment technologies for odor

removal should be considered like the super adiabatic combustor [68], water-jet plasma

scrubber [69] and externally oscillated burner [70] which were developed by the author. This

application in the technology should be expected that the system efficiency will be better than

conventional burning technologies. In addition, the order removal can be done by the

adsorption technology using the sludge char produced from the combined carbonization-

activator.

2.5 Summary

A novel rotary drum dryer was developed for the best drying of dewater sludge to produce

dried sludge which will be used for thermal treatment of the carbonization and steam

activation.

http://endic.naver.com/popManager.nhn?m=search&query=temporarily

29

The developed dryer was a new design, particularly in regards to the rotary kiln body

(deflector, pickup flights, internal screw vane, cylinder core) and inside rotating body (knife-

like blades, fork-like stirrers, fan-like blades). The newly designed parts can improve the

drying efficiency and the thermal efficiency, with lower volatile compounds in the dried

sludge, compared to conventional other dryers.

For verifying the effectiveness of this sludge drying, parametric screening studies were

conducted with varying the rotating drum temperature, the sludge residence time, and the

dryer load; the drying efficiency increased with increasing the rotating drum temperature and

the sludge residence time, while the efficiency decreased with increasing the dryer load.

Particularly, the rotating drum temperature may be a key factor to control the moisture

content in drying sludge. In addition, it was shown that NH3 and CO2 were the primary

components released from the sewage sludge drying process. The amounts of both

components increased when the rotating drum temperature increased.

For using the dried sludge for thermal treatments, the novel dryer should be considered

wholly in regards to a high drying efficiency, a high thermal efficiency, and lower volatile

compounds (particularly light organic compounds such as CH4). To satisfy these conditions,

the moisture content should be close to 10%. Therefore, the best operating conditions for the

developed dryer were set at 255℃ of the rotating drum temperature, 17 min of the sludge

residence time, and 55 kg/m3h of the dryer load. The average diameter of the dried sludge

created was about 8 mm and the weight reduction was 80%. The drying efficiency and the

moisture content in the dried sludge were 84.8 and 12.4%, respectively. And the thermal

efficiency was 73.8%, and the specific energy consumption was 3.49 MJ/kg of water which is

mostly the lowest value compared to other typical dryers.

30

Chapter 3

Pyrolysis and gasification performances

of the dried sludge Technologies for the thermal treatment of sewage sludge are appraised with reference to their

efficacy in terms of (a) operational parameters, (b) pre- and post-treatment requirements, and

(c) the extent of their use for the application. Particular attention is given to the characteristics

of the solid and gaseous products including generation which damages downstream

applications.

To produce energy (producer gas) and resource (sludge char) from the dried sewage sludge, a

study was achieved in three cases (the pyrolysis, the steam gasification, and the

carbonization-activation). Experiment was conducted with a batch-type wire-mesh reactor to

know the best method among above three cases for production of high quality sludge char and

producer gas, simultaneously. Comparative analysis on the formation characteristics of

products such as gas, tar, and char were evaluated for these cases.

It is effective to have carbonization-activation with the formation of volatile organic matter

through primary pyrolysis and steam activation during the secondary gasification process for

increasing the porosity of sludge char. The producer gas and sludge char which are produced

by the carbonization-activation should be utilized for renewable energy and resources.

The results through the performance in a fixed bed wire-mesh reactor will be used for

analyzing a pilot scale in Chapter 6.


The management of sewage sludge in an economically and environmentally acceptable

manner is one of the critical issues facing the society today. In fact, the amount of sewage

sludge produced by waste water treatment plants is going to dramatically increase in both

industrialized and emerging countries.

One of the main characteristics of the sewage sludge is the presence of high amounts of

inorganic ash and low carbon contents when compared with other materials, such as wood or

lignocellulosic residue from agriculture. As a result, sewage sludge has a relatively low

energy value, but is sufficient for some kind of waste-to-energy processes to be considered as

feasible. In addition, large amounts of sewage sludge are generated in every waste water

treatment plants, and as mentioned before, appropriate disposal methods need to be found

[71].

Recycling to agriculture (landspreading), incineration or landfilling is the most common

disposal routes. However, landspreading leads to an increase in concentration of heavy metals

in the soils and indirect emissions into air and water. Disposal by landfilling requires a lot of

space and poses a potential environmental hazard [72]. Incineration has the benefit of energy

recovery and volume reduction by 90%, but can induce secondary contamination due to high

concentration of heavy metal and air pollutants such as dioxin, SOx, NOx, etc.

Therefore, energy and/or resource utilizations of sewage sludge through the pyrolysis and

gasification into gas [73, 74], oil [75, 76], char [77-79], etc have been received the attention.

31

The pyrolysis and gasification can produce synthesis gas such as hydrogen, carbon monoxide,

methane, etc. and flammable hydrogen-rich gas. Oil from the pyrolysis and gasification

process can be recovered as biooil after purification. Heavy metals except mercury and

cadmium in sewage sludge can be stabilized inside of char, and they are not discharged [72].

However, liquefaction of biomass requires tremendous process cost and usage of high

pressure hydrogen during the separation of product from solvent [6]. In addition, high

concentration of nitrogen and sulfur inside of product hinders the utilization as fuel oil

without the additional process technology [80].

The pyrolysis is a process for producing gas or oil by high temperature thermal cracking via

external heat source without the supply of air or steam. Tar and refractory soot formation from

the pyrolysis process might cause the damage in subsidiary equipment. In addition, the

conventional gasification technology utilizes the partial combustion by minimizing the

amount of air to convert hydrocarbon into carbon monoxide, carbon dioxide, and hydrogen.

The formed synthesis gas has carbon dioxide and nitrogen, and those reduce the heating value.

But the steam gasification is known to produce syngas with higher concentration of hydrogen

from hydrocarbon component. Therefore, steam instead of air can be applied for gasification

to improve hydrogen content. In addition, steam will improve the specific area of char so that

it can be utilized as adsorbent.

Therefore, researches on the characteristics of the pyrolysis and the steam gasification for

energy and resource utilization are needed. Particularly, the investigation of producer gas,

char and tar which give damages to a machinery, is important to produce high quality product.

In this study, researches on the pyrolysis, the steam gasification, and the carbonization-

activation (i.e., sequential carbonization and steam activation) were conducted on dried

sewage sludge. The characteristics of each case were investigated. And physical properties of

sludge char, composition of producer gas and tar decomposition were evaluated.


3.2.1 Experimental setup Figure 3.1describes the lab-scale pyrolysis-gasification equipment which is batch type. It was

composed of a wire-mesh reactor, a steam-gas feeding line, a sampling-analysis line, and a

control-monitoring device.

The wire-mesh reactor was made of a cylindrical stainless pipe (500 mm in length and 85 mm

in inner diameter), and honeycomb-styled distributer was installed at the bottom section of a

reactor for uniform gas flow. For heating of the reactor, a micro-controlled electric furnace

(Model CLF-T1320, SERIN, Korea), which can heat up to 1000℃, was set around the reactor

wall. A cylindrical container, which was made by 400 mesh stainless steel wire matrix, was

positioned on top of the distributer. The cylindrical container was used for placing the dried

sludge sample.

The steam-gas feeding line was composed of a steam generator, a syringe water pump, and a

nitrogen cylinder. A steam generator with a cartridge heater in stick shape, generated steam by

controlling the temperature using a controller. The setting temperature was 350℃. Water

supplied by a syringe water pump (Model KDS 100, KD Scientific, USA) and the carrier gas

supplied from the nitrogen cylinder were introduced to the steam generator by passing the

venture mixer.

32

The sampling-analysis line was installed with wet and dry types sampling trap for tar

measurements during the pyrolysis-gasification, and with a wet type gas meter (Model W-

MK-10-ST, Shinagawa, Japan) to check the gas amount. In addition, two gas chromatographs

(Model 14B, SHIMAZU, Japan) were used for light tar and producer gas analysis. To protect

the gas chromatography column from the remaining tar and VOCs, the reformed gas passed

through the cotton and active carbon filters consecutively.

The control-monitoring device is for controlling and monitoring of temperatures in each part.

An electric furnace was operated by the heat controller to control the increasing rate of the

reactor temperature. A steam generator was controlled by a heat controller in order to operate

the setting temperature. In addition, thermocouples were installed at the sludge bed and the

producer gas in the reactor, on the wall of electric furnace, in the steam generator, etc. And a

data logger (Model Hydra data logger 2625A, Fluke, USA) was adopted for continued

monitoring of the temperatures in each part.

Figure 3.1 Experimental setup for pyrolysis-gasification tests

3.2.2 Experimental procedure

Sewage sludge used in this study had 81.3% of moisture content after dewatering by a

centrifuge at a municipal waste water treatment plant. To produce dried sludge within 10% of

moisture content, it was dried at 105~110℃ for about 7 hours using an electric furnace

(Model KS-35, Kwang Sung Co. Ltd, Korea).

The dried sludge was crushed by a grinder, and sieved using a Taylor sieve (Ro-Tap Sieve

Shaker, Chunggye Ltd., Korea) to 1~1.5 mm for having uniform diameter.

Table 3.1 shows proximate and ultimate analysis results of the dewatered sludge and the dried

sludge. The dried sludge contains 51.1% of volatile matter, and the elemental analysis showed

52.3% of carbon and 32.2% of oxygen, respectively.

Experiments were conducted for 3 cases, i.e. the pyrolysis, the steam gasification, and the

carbonization-activation. The temperature in a reactor, steam injection points and sampling

points were shown in Figure 3.2.

The dry sludge sample of 20 g was collected in a cylindrical container, and it was flushed on

<Section-detail drawing

for wire-mesh reactor>

33

top of the container at inner part of the reactor. By passing 100 mL/min of N2 gas as a carrier

gas for 30 minutes, purging was sufficiently made over a reactor and sampling line.

Table 3.1 Characteristics of dewatered sludge and dried sludge

Analysis method Contents Sewage sludge

Dewatered sludge Dried sludge

Proximate analysis (%)

Moisture 81.3 9.6

Volatile matter 12.5 51.1

Fixed carbon 1.5 6.4

Ash 4.7 32.9

Ultimate analysis (%)

C 46.9 52.3

H 6.8 8.2

O 33.6 32.2

N 8.4 7.92

S 4.3 0.01

The experiments for the pyrolysis and the steam gasification were conducted by heating up to

900℃ with 25 ℃/min of the heating rate without steam feed, and then the corresponding

temperature was maintained for 30 minutes. In the case of the pyrolysis, water was not fed.

But in the steam gasification, water is supplied at the point of 10 minutes (200℃). The water

flow rate was 30 mL/h.

In the carbonization-activation experiment, the reactor was heated up to 500℃ with the

heating rate of 25 ℃/min, and the holding time of 10 minutes was given. After that, steam of

30 mL/h was supplied, while it was heated up to 900℃, maintaining for 20 minutes after the

reaching this temperature.

To compare the results from the pyrolysis, the steam gasification, and the carbonization-

activation, total experiment time was fixed to 110 minutes for the all the cases.

After finishing each test, the carrier gas was supplied until the temperature inside of the

reactor was returned to the room temperature. After that, the top portion of the reactor was

opened to collect residual sludge char from the container for physical property evaluation.

(a) Pyrolysis and steam gasification (b) Carbonization-activation

Figure 3.2 Reactor temperature and sampling points for three cases

Elapsed time (min)

Re

acto

rte

mp

era

ture

(oC

)

0 10 20 30 40 50 60 70 80 90 100 1100

100

200

300

400

500

600

700

800

900

1000

150OC

350OC

Starting pointfor each case test

650OC

700OC

800OC

850OC

900OC

500OC

Sampling points

Steam injection pointfor steam gasification

750OC

Heating rate: 25oC/min

Holding time: 30min

Elapsed time (min)

Re

acto

rte

mp

era

ture

(oC

)

0 10 20 30 40 50 60 70 80 90 100 1100

100

200

300

400

500

600

700

800

900

1000

150OC

350OC

Starting pointfor each case test

560OC

700OC

800OC

850OC

900OC

500OC


Holding time: 20min

Sampling points


Steam injection pointfor carbonization-activation

Holding time: 10min

ActivationCarbonization

34

3.2.3 Sampling and analysis method for products

1) Tar sampling and analysis

To measure the total tar in the producer gas from the pyrolysis and/or gasification, the wet

type gravimetric tar mass was measured, and the concentration of light tar was determined

from the wet and dry type sampling simultaneously. Especially, the dry type sampling was

utilized for the analysis of formation and decomposition of light tar for elapsed time.

Light tars selected in this study were benzene, naphthalene, anthracene, and pyrene, which are

the representative ones as 1 to 4 benzene rings. In addition, nitrogen contained light tars were

selected as benzonitrile and benzeneacetonitrile [4].

Tar sampling lines for the wet type and dry type sampling were shown in Figure3.1.

The wet type tar sampling and analysis methods were according to the biomass technology

groups (BTGs) [81-85]. The wet type tar sampling was conducted over the total test time, and

tar was collected for weight and compositional analysis.

The wet type sampling train was consisted of two isothermal bath and ice bath with the

corresponding six impingers (200 mL) for condensation and absorption of tar. The first

isothermal bath was filled with water at the temperature lower than 20℃, and 100 mL of

isopropanol was filled in 4 impingers. The second ice bath was filled with isopropanol kept at

-20℃ using a chiller (ECS-30SS, Eyela Co., Japan), and one of two impingers was filled

with isopropanol. The other was left as empty.

In the series of impinger bottles, the first impinger bottle acts as a moisture and particle

collector, in which water, tar and soot are condensed from the producer gas by absorption in

isopropanol. Other impinger bottles collect tars, and the empty bottle collects drop.

Immediately after completing the sampling, the content of the impinger bottles were filtered

through a filter paper (Model F-5B, Advantec Co., Japan). The filtered isopropanol solution

was divided into two parts; the first was used to determine the gravimetric tar mass by means

of solvent distillation and evaporation by a evaporator (Model N-1000-SW, Eyela, Japan) in

which temperature and vapor pressure were 55~57℃ and 230 hPa, respectively. The second

was used to determine the concentrations of light tar compounds using a GC-FID (Model 14B,

Shimadzu, Japan).

The dry type sampling method was simpler compared with wet type method, and features

short sampling time in room temperature [81, 85]. Therefore, the dry type sampling method

was adopted to measure the time change of tar formation.

The dry type sampling method was employed using commercially available charcoal tubes

(Model 080150-054, Sibata Ltd, Japan) and silica gel tubes (Model 080150-061, Sibata Ltd,

Japan) which can collect tar by condensing and adsorbing tar in the producer gas.

The charcoal tube has 2 layers of activated carbon filled in glass tube of 6 mm outer diameter

whose weights are 100 mg and 50 mg, respectively. The silica gel tube also has 2 layers of

silica, 520 mg and 260 mg, filled in 8 mm outer diameter tube. Non-polar organic material

can be adsorbed well in the charcoal tube, and the silica gel tube can collect polar organic

material. Therefore, direct connection of charcoal and silica gel tube can achieve polar and

non-polar organic sampling simultaneously [81].

The flow rate for the dry sampling was set to 0.15 L/min for 5 minutes. After finishing the

sampling, adsorbed tar was stored in a refrigerator to prevent evaporation. Carbon disulfide of

35

2 mL dissolves adsorbed tar in the charcoal tube. Acetone of 4 mL does the same for tar in the

silica gel tube. Dissolved tar solution was centrifuged for 2 h for complete dissolution of tar

from charcoal and silica gel (Model TTS 2. IKA Works Inc., USA).

Quantitative tar analysis was performed by a GC system, using a RTX-5 (RESTEK) capillary

column (30 m - 0.53 mm id, 0.5 μm film thickness) and an isothermal temperature profile at

45℃ for the first 2 minutes, followed by 7 ℃/min temperature gradient to 320℃ and finally

an isothermal period at 320℃ for 10 minutes. Helium was used as a carrier gas. The

temperature of the detector and the injector were maintained at 340℃ and 250℃ ,

respectively.

2) Sampling and analysis for producer gas

Producer gas was sampled at the downstream of the wet type tar sampling trap as can be seen

in Figure 3.1. A set of backup VOC adsorber was installed downstream of the series of

impinge bottles to protect the column of the gas chromatography from the residual solvent or

VOCs, which may have passed through the impinger train. The set of backup VOC adsorber

consisted of two cotton filters and an activated carbon filter connected in series. Gas analysis

was conducted using a GC-TCD (Model CP-4900, Varian, Netherland). For the measurement

of H2, CO, O2, and N2, 5A of molecular sieve column was used. PoraPlot-Q column was used

for CO2, C2H4, and C2H6 to achieve simultaneous analysis.

3) Sludge char analysis

To determine pore development of sludge char, the nitrogen adsorption test was conducted.

Using nano POROSITY (Model NanoPOROSITY-XQ, MiraeSI, Korea), N2 adsorption ability

was measured by taking the isothermal adsorption curve at -196℃ [71]. According to this,

adsorption characteristics were analyzed and specific area was calculated by using a BET

equation.

Determination of the pore distribution and average pore size was conducted by the HK

(Horvath–Kawazoe) and BJH (Barret-Joyner-Halenda) equations for micropore and mezopore,

respectively.

To compare pore development in the sludge char, SEM (scanning electron microscopy; Model

S-4800, Hitachi Co., Japan) was used at 50,000 times of resolution. Chemical characteristics

and composition were analyzed using an EDX (energy-dispersive X-ray spectroscopy; Model

7593-H, Horiba, UK).

In addition, benzene adsorption test was used for evaluating tar adsorption characteristics in

the sludge char.

3.2.4 Test setup and procedure for benzene adsorption

Benzene adsorption test was conducted by using a test setup of shown in Figure 3.3. N2 was

fed to the benzene feeding line (17.5 mL/min) and to the dilution line (1.98 L/min) through

the isothermal bath maintained at the temperature of 25℃. Benzene gas mixture passed a U

tube filled with a sludge char sample of 20 grams. 1 ml of gas sampling was conducted by a

syringe (Model 22265, SUPELCO, USA) at the inlet and outlet of the U tube for 5 minutes

36

interval. At this time, the sludge char in a U tube was also weighted for adsorption amount.

Quantitative benzene analysis was performed by a GC-FID (Model GC-14B, SHIMADZU,

Japan), using a RTX-5 (RESTEK) capillary column (30 m - 0.53 mm id, 0.5 μm film

thickness) and an isothermal temperature profile at 45℃ for the first 2 minutes, followed by a

7 ℃/min temperature gradient to 100℃ and finally an isothermal period at 100℃ for 2

minutes. Helium was used as a carrier gas. The temperature of the detector and the injector

were maintained at 340℃ and 250℃, respectively.

Figure 3.3 Test equipment for benzene adsorption of the sludge char


3.3.1 Effects of pyrolysis, gasification and carbonization-activation

As for the pyrolysis process, TGA (thermo-gravimetric analysis) and DTG (derived thermo-

gravimetric) analysis were conducted to clarify the relationship between the weight loss and

the pyrolysis temperature, and the results were shown in Figure 3.4 [72]. The test was

conducted using the dewatered sludge (Table 3.1), and the heating rate was 25 ℃/min.

The pyrolysis of the dried sludge could be elucidated by two steps after moisture evaporation

in 100~150℃, as shown in a DTG curve. First step was related to volatile component in the

range of 200~500℃, and the second one was decomposition of inorganic compound over

500℃.

The first step features two peaks, and it can be explained as the following. The first peak

might be due to decomposition and devolatilization of less complex organic structures which

is a small fraction. The second peak was caused by decomposition of more complex organic

structures corresponding to a larger fraction. The second step was related to decomposition of

inorganic matter as described before.

In the first step, TG displayed 57% at 500℃, which can be explained as 43% weight loss from

the initial state (i.e., the total weight). The second step was found to be 46.2% at 900℃, which

shows 53.8% loss of the total weight. The difference between both steps was 10.8%. This is

because the second step was corresponded to decomposition of an inorganic matter which has

physical property of lower reduction in weight.

37

Figure 3.4 TGA and DTG for the pyrolysis of the dried sewage sludge

As explained in the Figure 3.4, the pyrolysis process is a thremo-chemical decomposition in

the absence of oxidizing agent like steam, CO2, air, etc.

The steam gasification is a process that converts sludge to product with steam agent. The

steam injects to a reactor in the beginning of the process, giving predominated reactions such

as water gas reaction (Eq. 3.2), tar steam gasification (Eq. 3.6) and water-gas shift reaction

(Eq. 3.8).

The carbonization-activation is the process having two steps; sequential pyrolysis and steam

gasification. But the second step is fed small amount of the steam for activation only making

micro-pores, compared to conventional steam gasification.

The pyrolysis and gasification of the dried sludge can be classified as primary pyrolysis and

the secondary reaction (i.e., the pyrolysis and/or the steam gasification) as shown in Figure

3.5. The dried sludge was converted into char, tar, and gas during the primary pyrolysis

process, and it was further converted into gas from tar and char during the secondary reaction.

Sewage sludge is mainly composed of cellulose, and the significant amount converted to tar

during the primary pyrolysis. Then, some part of the tar was known be change to gas at the

secondary reaction [86].

Figure 3.5 Material balance for the pyrolysis and the steam gasification

Temperature (oC)

TG

(wt%

)

DT

G(w

t%s

-1)

0 100 200 300 400 500 600 700 800 9000

20

40

60

80

100

0

0.05

0.1

0.15

0.2

38

The primary pyrolysis was mainly affected by the heating rate, and the secondary reaction

was determined by the reactor temperature [87, 88]. The secondary pyrolysis and/or

gasification can be the one of the followings: char gasification reactions (Eqs. 3.1~3.4); tar

decomposition reactions (Eqs.3.5~3.6); light gas reactions (Eqs. 3.7~3.10), etc.

▌Char gasification reaction

▰ Partial oxidation

C + 1/2O2 → CO ΔH = - 110.5 kJ/mol (3.1)

▰ Water gas reaction

C + H2O → CO + H2 ΔH = 131.3 kJ/mol (3.2)

▰ Boudouard reaction

C + CO2 → 2CO ΔH = 171.7 kJ/mol (3.3)

▰ Hydrogasification

C + 2H2 → CH4 ΔH = - 74.9 kJ/mol (3.4)

▌Tar decomposition reaction

▰ Tar pyrolysis

Tar → wH2 + xCO + yCO2 + zCnHm (3.5)

▰ Tar steam gasification

Tar + vH2O → xCO + yH2 (3.6)

▌Light gas reaction

▰ Methanation reaction

CO + 3H2⇌ CH4 + H2O ΔH = - 206.2 kJ/mol (3.7)

▰ Water-gas shift reaction

CO + H2O ⇌ CO2 + H2 ΔH = - 41.1 kJ/mol (3.8)

In addition, high temperatures were also responsible for the reduction of C2H4 and C2H6.

Some of the typical reactions are as follows [89]:

C2H6 → C2H4 + H2 (3.9)

C2H4 → CH4 + C (3.10)

1) Mass yield in product

Figure 3.6 compared mass yield of char, tar, and gas from the pyrolysis, the steam gasification,

and the carbonization-activation, respectively. The char and the tar were measured

respectively, but the gas was calculated from the both value for three cases.

The pyrolysis without the steam feed formed 43.9% of sludge char, 22.3% of tar, and 33.8%

of producer gas. The total amount of the producer gas was 11.5 L. The volume of the sludge

char was reduced due to decomposition of organic structure in the primary pyrolysis, and

heterogeneous reaction of carbon and residual inorganic decomposition in the secondary

pyrolysis process. Tar was formed during the primary pyrolysis, and it was converted into

producer gas in the secondary pyrolysis. The producer gas was the total gas from the primary

and secondary pyrolysis processes.

39

The steam gasification was conducted by continuously feeding steam from the beginning of

the process. Product was 39.2% of sludge char, 23.5% of tar, and 37.3% of producer gas. The

total amount of the producer gas was 20.1 L.

The steam gasification showed 10.7% reduction of sludge char compared with the pyrolysis.

However, tar and gas displayed 5.38% and 10.35% increases, respectively.

The gas increases due to the water gas reaction (Eq. 3.2) and tar steam gasification (Eq. 3.6)

by the steam supply. Although some portion of tar converted to light gas, the tar increased due

to the steam effect. The role of steam cannot be limited only for transportation and

stabilization of the volatile products, such as nitrogen which was used as carrier gas in this

study. The ability of steam to penetrate into solid materials and to help desorption, distillation,

and efficient removal of the volatile products from it can explain the higher yield of tar [90].

The carbonization-activation was conducted by the pyrolysis up to 500℃ and then

gasification by feeding steam. The product was 40.1% of sludge char, 22.7% of tar, and

37.2% of gas. Total amount of producer gas was 16.5 L. For the carbonization-activation,

2.29% increase of sludge char, 3.4 and 0.26% reduction of tar and producer gas, respectively

were found compared with the steam gasification. This can be elucidated by less chance of

water gas reaction (Eq. 3.2) due to steam feeding after carbonization. In addition, relatively

small amount of gas reduction than tar was induced by the conversion into gas from tar due to

tar steam gasification (Eq. 3.6).

Figure 3.6 Mass yield of product and total gas amount

2) Characteristics of producer gas Figure 3.7 shows the cumulative gas amount and the instantaneous gas amount.

The pyrolysis, the steam gasification, and the carbonization-activation displayed the increased

cumulative gas amount according to the elapsed time. The order of increment ratio was the

steam gasification, the carbonization-activation, and the pyrolysis. As discussed before, it

could be elucidated by the increase in converted gas amount from char and tar according to

water gas reaction (Eq. 3.2) and tar steam gasification (Eq. 3.6).

Ma

ss

yie

ld(%

)

To

talg

as

am

ou

nt

(L)

0

10

20

30

40

50

0

10

20

30

Total gas amountChar Tar

Steam gasification

Pyrolysis

Carbonization-activation

11.5

16.5

20.1

43.9

22.3

23.5

40.1

37.2

39.2

22.7

37.3

33.8

Gas

40

The instantaneous gas amount displayed the primary peak for three cases. This was found to

be related to evaporation of the volatile matter during the primary pyrolysis process as

explained by the DTG curve shown in Figure 3.4. And although the steam gasification fed

water from 200℃, the steam did not affect for producing gas.

After this region, the steam gasification and the carbonization-activation was significantly

increased according to water gas reaction (Eq. 3.2) and tar steam gasification (Eq. 3.6). In this

case, the carbonization-activation showed fewer amount than the steam gasification due to

delayed steam reaction. In addition, regardless of the steam effects, the secondary peak was

caused by the sludge char gasification (Eqs. 3.1, 3.3, 3.4) and tar pyrolysis (Eq. 3.5) at this

high temperature reaction zone.

(a) Pyrolysis (b) Steam gasification (c) Carbonization-activation

Figure 3.7 Gas production amount according to the elapsed time

Figure 3.8 shows the gas yield and energy yields of each constituent of the light gases from

the pyrolysis, the steam gasification, and the carbonization-activation. The gas yield was

calculated by each gas concentration and the total gas amount, and the energy yield was taken

by higher heating value and the gas yield.

For the pyrolysis, the producer gas was found to be H2, CO, CO2, CH4, C2H4, C2H6 according

to the amount order. The energy yield was 109 kJ. As already explained in Figures 3.3 and 3.4,

the light gases generated from primary decomposition of less complex organic structures (i.e.,

CO, CO2) and from secondary char gasification, tar decomposition, light gas reactions.

Figure 3.8 Gas and energy yields of each case

Elapsed time (min)

Insta

nta

ne

ou

sg

as

am

ou

nt

(m3/h

-kg

_slu

dg

e)

Cu

mu

lative

ga

sa

mo

un

t(L

)

0 20 40 60 80 1000

0.3

0.6

0.9

1.2

0

10

20

30

40

50

Cumulative gas amount

Instantaneous gas amount

Elapsed time (min)

Insta

nta

ne

ou

sg

as

am

ou

nt

(m3/h

-kg

_slu

dg

e)

Cu

mu

lative

ga

sa

mo

un

t(L

)

0 20 40 60 80 1000

0.3

0.6

0.9

1.2

0

10

20

30

40

50



Steam injection point

Elapsed time (min)

Insta

nta

ne

ou

sg

as

am

ou

nt

(m3/h

-kg

_slu

dg

e)

Cu

mu

lative

ga

sa

mo

un

t(L

)

0 20 40 60 80 1000

0.3

0.6

0.9

1.2

0

10

20

30

40

50




Ga

syie

ld(L

)

En

erg

yyie

ld(k

J)

0

2

4

6

8

10

0

100

200

300

400

H2

CO CO2

CH4

C2H

4 Energyyield

Pyrolysis

Steam gasification


3.75

9.87

7.85

3.34

2.86

109

0.12 0.11

5.07

3.53

291

0.19 0.18

4.65

2.71

1.13

0.12

226

0.05

1.23

1.35

C2H

6

41

For the steam gasification, the order of gas yield was similar to the pyrolysis process, but the

amount was larger than the pyrolysis. Especially, significant increases in hydrogen and carbon

monoxide were found compared with the pyrolysis. This could be explained by the

introduction of steam for reforming reactions, such as water gas reaction (Eq. 3.2), tar steam

gasification (Eq. 3.6), and water-gas shift reaction (Eq. 3.8). The energy yield was also

increased to 291 kJ.

For the carbonization-activation, the gas yield was also similar to the steam gasification, but it

was slightly reduced along with the energy yield, 226 kJ. The reason is that the carbonization-

activation had lower chance due to delayed steam feed, compared to the steam gasification.

3) Characteristics of tar formation

Figure 3.9 describes the concentration of light tar according to the elapsed time for the

pyrolysis, the steam gasification, and the carbonization-activation. The dry type sampling

method was adopted to display selected light tar concentrations in given time.

The pyrolysis displayed the maximum around 700℃ as shown in Figure 3.9(a). This is

because that the heavy tar from the primary pyrolysis was converted into light tar (light

aromatic tar and light PAH) during the secondary pyrolysis. Each maximum concentration of

the light tar was 11.2392 g/Nm3 for benzene (1 ring), 1.0519 g/Nm

3 for naphthalene (2 rings),

0.0738 g/Nm3 for anthracene (3 rings), 0.0124 g/Nm

3 for pyrene (4 rings), 0.3411 g/Nm

3 for

benzonitrile (1 ring), and 0.3545 g/Nm3

for benzeneacetonitrile (1 ring). After displaying the

maximum, the selected tars were decreased due to conversion of the light tar into light

hydrocarbons as described in Figure 3.4 [15]. Benzonitrile and benzeneacetonitrile was

known to be representative material in nitrogen-containing tar during sludge pyrolysis [4].

For the steam gasification, similar pattern was exhibited as the pyrolysis process, but the

maximum amount was higher than the pyrolysis process as shown in Figure 3.9(b). The steam

penetration into solid materials helps efficient removal of the volatile products which includes

heavy and light tars [90]. Also, some part of heavy tar converted to light tar. This is due to

largest amount of light tar at the primary pyrolysis according to the steam injection into

reactor at 200℃. Each maximum concentration was 16.1466 g/Nm3 for benzene, 1.7823

g/Nm3 for naphthalene, 0.1733 g/Nm

3 for anthracene, 0.0171 g/Nm

3 for pyrene, 1.497 g/Nm

3

for benzonitrile, and 0.965 g/Nm3 for benzeneacetonitrile.

For the carbonization-activation, different characteristics from the pyrolysis and the steam

gasification were observed. Small amount of tar was formed up to 500℃ similar to the

pyrolysis, which was steam injection temperature. However, after the steam injection, tar

destruction by the tar steam gasification (Eq. 3.6) was preferred instead of the formation of

light tar like Figure 3.9(a). Each maximum concentration was 5.8103 g/Nm3 for benzene,

0.4347 g/Nm3 for naphthalene, 0.0756 g/Nm


3 for pyrene, 0.2983

g/Nm3 for benzonitrile, and 0.2624 g/Nm

3 for benzeneacetonitrile.

42


Figure 3.9 Tar generation in each case according to the elapsed time

Figure 3.10 shows the result of analyzed concentration to compare the formation

characteristics of light tar from the pyrolysis, the steam gasification, and the carbonization-

activation. The light tars were measured by the wet type sampling method for the same test

time of 65 minutes.

The amount order of light tar showed the same pattern for the three cases as benzene,

naphthalene, benzonitrile, benzeneacetonitrile, anthracene, and pyrene.

For the pyrolysis, light aromatic tar (i.e., benzene) was formed significantly compared with

light PAHs (i.e., naphthalene, anthracene, and pyrene). This could be elucidated by the

significant conversion of light aromatic tar from light PAH due to the thermal cracking, etc. In

addition, the steam gasification showed reduced amount due to conversion into light aromatic

tar and producer gas from light PAH through steam supply. However, the carbonization-

activation displayed fewer amount than the other two cases. As discussed in Figure 3.8, most

of the heavy tar remained did not convert into light tar in the primary pyrolysis, and then the

heavy tar converted to small amount of light tar after feeding steam for the secondary

gasification.

Figure 3.10 Light tar contribution for each case

Elapsed time (min)

Ta

rco

nce

ntr

atio

n(g

/N

m3)

Re

acto

rte

mp

era

ture

(oC

)

0 20 40 60 80 1000

4

8

12

16

20

0

200

400

600

800

1000

Temperature

Anthracene

Benzene

Benzen-acetonitrile

Naphthalene

Pyrene

Benzonitrile

Elapsed time (min)

Ta

rco

nce

ntr

atio

n(g

/N

m3)

Re

acto

rte

mp

era

ture

(oC

)

0 20 40 60 80 1000

4

8

12

16

20

0

200

400

600

800

1000

Temperature

Anthracene

Benzene

Benzen-acetonitrile

Naphthalene

Pyrene

Benzonitrile


Elapsed time (min)

Ta

rco

nce

ntr

atio

n(g

/N

m3)

Re

acto

rte

mp

era

ture

(oC

)

0 20 40 60 80 1000

4

8

12

16

20

0

200

400

600

800

1000

Temperature

Benzene

Benzonitrile

Naphthalene

Benzene-acetonitrile

Pyrene

Anthracene


Ta

rco

nce

ntr

atio

n(g

/N

m3)

0

1

2

3

4

5

6

Pyrolysis

Steam gasification


Benzene Naph-talene

Anthra-cene

Pyrene Benzo-nitrile


5.5

1.36

0.35 0.11

0.530.23

4.41

1.11

0.23 0.140.43

0.25

2.79

0.75

0.14 0.140.38

0.17

43

For the three cases, nitrogen-containing tars (i.e., benzonitrile and benzeneacetonitrile)

showed the similar phenomena with light tar, but small amount of nitrogen in sludge (refer

Table 3.1) forced the less formation than benzene (light aromatic tar). The nitrogen-containing

tar can cause environmental hazard due to the conversion into ammonia (NH3) and hydrogen

cyanide (HCN) during the pyrolysis, which are precursor of NOx [91].

4) Characteristics of sludge char

▌Physical properties of sludge char Figure 3.11 compares the incremental pore volume and SEM micrographs to determine pore

characteristics of the sludge char.

The pore size classification in this study follows the IUPAC classification [63, 92] i.e.

micropores (<2 nm), mesopores (2~100 nm) and macropores (>100 nm).

For the pyrolysis, macropore was distributed. However, the steam gasification and the

carbonization-activation displayed uniform distribution from micropore to macropore. It

could be confirmed in SEM at 50,000X resolution. Especially, the carbonization-activation

showed the highest overall cumulative pore volume.

For the carbonization-activation, pores were developed from evaporation of moisture content

and volatile matter during the primary pyrolysis (i.e., carbonization). Bagreev et al. proved

that water released by the dehydroxylation of inorganic material could aid pore formation and

moreover could act as an agent for creating micropores [93]. In addition, Inguanzo et al.

propose that carbonization increased the porosity through unblocking many of the pores

obscured by volatile matter [94]. The secondary steam gasification process (i.e., the steam

activation) would develop uniform pore along with gas formation through on-site surface

reaction after penetration of hot steam into the existing pore, which was injected from outside.


Figure 3.11 Incremental pore volume and SEM photos for sludge chars in each case

Table 3.2 describes pore characteristics of the sludge char. As expected from Figure 3.11, the

carbonization-activation showed the lowest mean pore size with the maximum specific

surface area and pore volume, compared to other two cases.

Pore width (nm)

Incre

me

nta

lp

ore

vo

lum

e(c

m3/g

)

0 50 100 150 2000

0.0005

0.001

0.0015

0.002

0.0025

0.003

2 nm

Pore width (nm)

Incre

me

nta

lp

ore

vo

lum

e(c

m3/g

)

0 50 100 150 2000

0.0005

0.001

0.0015

0.002

0.0025

0.003

2 nm

Pore width (nm)

Incre

me

nta

lp

ore

vo

lum

e(c

m3/g

)

0 50 100 150 2000

0.0005

0.001

0.0015

0.002

0.0025

0.003

2 nm

44

Table 3.2 Porous characteristics of the sludge char for each case

Condition Mean pore size

(nm)

Specific surface area

(m2/g)

Pore volume

(cm3/g)

Pyrolysis 11.104 14.9 0.0414

Steam gasification 6.229 77.6 0.1203

Carbonization-

activation 6.203 80.3 0.1351

Table 3.3 represents the analysis and test results of the commercial activated carbon of which

benzene adsorption test were conducted in this study. In addition, two adsorbents (activated

carbon and wood chip) as studied by Phuphuakrat et al. were shown [81].

As for the sludge char produced from the carbonization-activation process, its mean pore size

was bigger than those of commercial activated carbon and activated carbon, and its specific

surface area and pore volume were smaller. The sludge char (6.203 nm) had a mean pore size

of mesopores, like wood chips (10.077 nm), but two activated carbon materials had a mean

pore size of micropores. That is, the sludge char and wood chips had the characteristic similar

to adsorption of mesopores. And the sludge char had a specific surface area of 80.3 m2/g, an

influential factor for adsorption amount, larger than that of wood chips (1.1 m2/g).

In the case of the adsorption amount, the sludge char (174.8 mg/g) and commercial activated

carbon (586.2 mg/g) were used for the benzene test in this study. But the activated carbon and

wood chips were used by Phuphuakrat et al. for the total tar adsorption measurement from the

pyrolysis of Japanese cedar. Therefore, it is difficult to compare the adsorption capacity

results from the two cases.

However, the benzene adsorption capacity of the sludge char was relatively smaller compared

to that of commercial activated carbon in this study. As mentioned above, the reason is that

the pore size was larger and the specific surface area was smaller in the sludge char.

Table 3.3 Porous characteristics and adsorption capacity of the adsorbents from this study and

other results [81]

Adsorbent Mean pore size

(nm)

Specific surface

Area(m2/g)

Pore volume

(cm3/g)

Adsorption

amount(mg/g)

Commercial

activated carbon1)

1.830 1376.6 0.6300 586.2

Activated carbon2)

1.128 987.1 0.5569 97.5

Wood chip2)

10.077 1.1 0.0058 155.7

<Note> 1)

Commercial activated carbon (tested by benzene adsorption in this study); 2)

Tested by Phuphuakrat et

al. (tar adsorption from the pyrolysis of Japanese cedar)

Some portion of heavy tar, that is almost all of the gravimetric tar shown in Figure 3.3, was

converted to light tar or gas via the thermal cracking or the steam reforming.

Among the light tars, non-condensible aromatic tars (e.g., benzene, benzonitrile,

benzoacetonitrile, etc.) will give no damage to devices (i.e., combustor, engine, pipe line, etc).

45

That is, when light aromatic tars are contained in the producer gas, they can be effectively

used for increasing the heating value and thus the energy efficiency. However, condensable

amounts of light PAH tars such as naphthalene, anthracene, and pyrene may give damages to

the devices and thus need to be removed.

Therefore, it should be effective that adsorption of the sludge char could not be better for the

non-condensible light aromatic tar to improve the energy efficiency. For the sludge char

produced from the steam activation, the benzene passed through the sludge char after it

reached the breakthrough point in a short time. As mentioned above in the tar adsorption test,

as determined by Phuphuakrat et al. [81], the development of mesopores in the sludge char

easily leads to the adsorption of condensable light PAH tar, making it an efficient adsorbent

for tar removal of producer gas generated from the pyrolysis and gasification.

Figure 3.12 shows the N2 adsorption-desorption isotherm for the sludge char of each case.

The analysis on the adsorption isotherm provides an assessment for the pore size distribution.

According to the isothermal adsorption graphs, the sludge char from the pyrolysis exhibited

only a small amount of adsorption, but the sludge chars from the steam gasification and the

carbonization-activation displayed a larger amount of adsorption. Especially, the sludge char

from the carbonization-activation showed the slightly larger amount of adsorption compared

with the steam gasification.

According to the IUPAC classification, the curve of the sludge char, particularly for the steam

gasification and the carbonization-activation, corresponds to Type V isotherm. A characteristic

of the Type V isotherm is the hysteresis loop, which is associated with the capillary

condensation in pores and limiting uptake at high relative pressure [95].


Figure 3.12 Isothermal adsorption linear plot

A semi quantitative chemical analysis of the sludge chars in each case, Figure 3.13 and Table

3.3, was obtained from the EDX analyzer coupled to SEM measurements.

Carbon component showed the fewer amounts for the steam gasification compared with the

pyrolysis case. It could be originated from the impact of water gas reaction (Eq. 3.2) due to

steam injection. For the carbonization-activation, larger amount of residual carbon was

displayed because of the delayed steam injection compared with the steam gasification.

Inorganic atoms might be considered as the potential catalysts for the pyrolysis or the steam

gasification reaction. For example, with Al, if existing in the form of Al2O3, it would be an

acid catalyst for the cracking reaction [96]; or with K and Ca atoms, they were already

reported as the catalyst for biomass pyrolysis in literature [97].

Relative pressure (P/Po)

Qu

an

tity

ad

so

rbe

d(c

m3/g

ST

P)

0 0.2 0.4 0.6 0.8 10

20

40

60

80

100

Desorption

Adsorption


Qu

an

tity

ad

so

rbe

d(c

m3/g

ST

P)

0 0.2 0.4 0.6 0.8 10

20

40

60

80

100

Desorption

Adsorption


Qu

an

tity

ad

so

rbe

d(c

m3/g

ST

P)

0 0.2 0.4 0.6 0.8 10

20

40

60

80

100

Desorption

Adsorption

46

Figure 3.13 Element compounds measured by EDX

Table 3.3 Content of elements in the sludge char for each case (wt.%)

C O Mg Al Si P S Cl K Ca Ti Fe Zn Ba

Pyrolysis 52.41 44.27 0.08 0.55 0.84 0.66 0 0.03 0.23 0.36 0 0.44 0.01 0.1

Steam

gasification 50.21 42.91 0.09 1.02 1.76 1.47 0.03 0 0.28 0.59 0.02 0.74 0.04 0.03

Carbonization

activation 51.08 43.78 0.09 0.74 1.21 0.8 0 0.02 0.24 0.57 0.06 0.56 0 0.03

▌Benzene adsorption properties

To evaluate adsorption properties of the sludge chars, which were produced by different

methods, the test setup (Figure 3.3) was used. The input concentration of benzene for the

adsorption test was fixed at 1 %. GHSV (gas hourly space velocity) at 25℃ was 6,315/h (Gas:

2 L/min, Volume: 19 mL).

Figure 3.14 shows a breakthrough curve, the adsorption amount and the reaching time to

saturation point for comparing adsorption characteristics of the sludge chars. C is effluent

concentration, and Ci refers the input concentration. Breakthrough point was defined to be

happened more than 10% for the ratio of effluent concentration and input.

For the pyrolysis, the breakthrough point appeared right after starting adsorption. Saturation

point was reached in 5 min. At this time, the adsorption amount was 18 mg/g. The steam

gasification displayed the breakthrough point at 10 minutes. The saturation point was 30

minutes while the adsorption amount was 157 mg/g. In case of the carbonization-activation,

the saturation point of the activated char displayed the longest time as 35 minutes, and the

adsorbed amount showed the largest value as 175 mg/g. This trend could be explained by

Figure 3.11 because the case of the carbonization-activation shows the most developed micro-

pores and mesopores.

Energy (keV)

Co

un

ts

1 2 3 4 5 6 70

200

400

600

Energy (keV)

Co

un

ts

2 4 6 8 100

200

400

600

Zn

CO

Mg

Al

Si


S Cl

K Ca

Ti Ba Fe

P

Pyrolysis

Steam gasification

C

O

Mg

Al

Si

P

S ClK Ca

Ti BaFe

47

Figure 3.14 Breakthrough curve and the adsorption amount of benzene for the sludge chars

3.3.2 Verification of adsorptive tar removal from a continuous pyrolyzer

1) Test setup and procedure for the sludge char adsorption

Figure 3.15 exhibits the test equipment diagram designed to verify tar adsorption performance

for the sludge char produced during the carbonization-activation in the wire-mesh reactor. A

screw pyrolyzer was used continuously to supply the tar-containing gas which generated by

wood chips pyrolysis.

The sludge char was sieved using a Taylor sieve (Ro-Tap Sieve Shaker, Chunggye Ltd., Korea)

to 1~1.5 mm for having uniform diameter. Sludge char of 40g (120 mL) in similar diameter

were fixed at the U-shaped absorber which maintains atmosphere temperature. And the tar

containing gas produced from the pyrolyzer was fed into it with the gas hourly space velocity

(GHSV) [98] of 400/h. The gas was cooled down to the temperature of 25~30℃ by natural

convection, without any heat exchanger device.

To evaluate tar adsorption capacity for the sludge char, the wet type gravimetric tar mass was

measured, and concentrations of light tars were determined from the wet and dry types

sampling simultaneously. Especially, the wet type light tar were measured by the group of

light tar components divided by each benzene ring [85]. And the dry type light tar was utilized

for showing the break through curves according to the elapsed time.

The wet type sampling was performed for an hour while the flow rate of the pyrolysis gas was

fixed at 0.8 L/min after a stabilization period. Another sampling followed for an hour from the

start of the adsorption process during which the pyrolysis gas was passing through the

absorber steadily.

The dry type sampling was performed to determine breakthrough curves. The first dry type

sampling was performed prior the adsorption to determine the input concentration of the

pyrolysis gas after a stabilization period. The flow rate of the pyrolysis gas was fixed at 0.5

L/min. The second session of the dry type sampling was conducted for one-and-a-half hours

with 3 minutes intervals for the first hour and 5 minutes intervals for the latter half hour while

Sa

tura

tio

np

oin

t(m

in)

0

10

20

30

40

Pyrolysis Steamgasification

Carbonzationactivation

Adsorption time(min)

C/C

i

Ad

so

rptio

na

mo

un

t(m

g/g

)

0 10 20 30 40 500

0.2

0.4

0.6

0.8

1

0

50

100

150

200

250

Pyrolysis

Steamgasification

Carbonzation-activation

48

the pyrolysis gas was passing through the absorber.

The tar analysis was conducted according to the method of biomass technology groups (BTGs)

[82]. The detailed tar analysis methods can be referred to “3.2.3 Sampling and analysis

method for products; 1) Tar sampling and analysis”.

Figure 3.15 Experimental setup for adsorption test in the biomass tar from a screw pyrolyzer

2) Adsorption characteristics of biomass tar

The experimental results of the adsorption process using sludge chars as an absorbent for

biomass tar are presented in Figures 3.16 and 3.17.

Figure 3.16 shows the gravimetric tar mass and the concentrations of representative light tar

components by individual and group [85] before and after adsorption. The gravimetric tar had

the concentration of 19.87 g/Nm3 at the inlet of the absorber but dropped to 6.82 g/Nm

3 at the

outlet, showing the adsorption efficiency of 65.7%. The moderate adsorption efficiency was

due to Group 1 of light aromatic tar which mostly remained unabsorbed.

Figure 3.16 Tar contribution before and after adsorption

Lig

ht

tar

(g/N

m3)

0

1

2

3

4

5

6

7

Group 4Group 1 Group 2 Group 3

Gra

vim

etr

icta

r(g

/Nm

3)

Lig

ht

tar

(g/N

m3)

0

5

10

15

20

0

1

2

3

4

Before adsorption

After adsorption

Naph-thalene

Anthra-cene



Gravimetrictar

Benzene

6.82

0.003

19.9

0.05

2.96

0.61

3.19

0.28

0.02

0.08

0.75

0.01

0.07

0.01

49

However, the tar concentrations by higher group 2 showed a significant reduction like

individual compounds although the concentration values were higher in certain groups. The

concentration dropped to 5.09 g/Nm3 from 5.93 g/Nm

3 for Group 1; 0.17 g/Nm

3 from 1.14

g/Nm3 for Group 2; 0.03 g/Nm

3 from 0.35 g/Nm

3 for Group 3; and 0.02 g/Nm

3 from 0.18

g/Nm3 for Group 4.

The concentrations of the light tar compounds were also reduced to 2.96 g/Nm3 from 3.19

g/Nm3 for benzene; 0.28 g/Nm

3 from 0.75 g/Nm

3 for naphthalene; 0.01 g/Nm

3 from 0.07

g/Nm3

for anthracene; 0.003 g/Nm3 from 0.05 g/Nm

3 for pyrene; 0.02 g/Nm

3 from 0.61

g/Nm3 for benzonitrile; and 0.01g/Nm

3 from 0.08g/Nm

3 for benzoacetonitrile after the

adsorption process.

Figure 3.17 shows the breakthrough curves of the sludge char for selected light tar

components during the adsorption process.

The benzene steadily adsorbed through the micropore in the sludge char up to 35 minutes

from the start of the adsorption process; however, it passed thereafter without any adsorption.

The naphthalene at the exit of the adsorption bed gradually increased, and the value of C/Ci

reached about 27% after passing 55 minutes from the start of the adsorption experiment. The

anthracene and pyrene at the exit of the adsorption bed gradually increased. The anthracene

value of C/Ci reached 10% after passing 35 minutes and the pyrene value of C/Ci reached 5%

after passing 50 minutes.

The breakthrough curves showed that the sludge char did not adsorb non-condensable tarlike

benzene (i.e., Group 1). Such non-condensable tar is easily adsorbed by micropore adsorbent

materials such as the activated carbon [99], while the pores of sludge char are in the range of

mesopore. Therefore, non-condensable tar can hardly be adsorbed by the sludge char. But the

condensable tars like Group 2 or heavier were relatively well removed in the sludge char.

Figure 3.17 Breakthrough curves of the sludge char for light tar adsorption

Heavy tar from the pyrolysis and/or gasification should be converted into light tar and light

gas through the thermal cracking and the reforming. If the light tar was fed into a combustor

Elasped time (min)

C/C

i

0 10 20 30 40 50 60 70 80 900

0.2

0.4

0.6

0.8

1

Benzene

Naphthalene

Anthracene

Pyrene

50

or an engine without condensation, there will be no damage in machinery and will result in

higher energy efficiency due to increased heating value. That is, the inclusion of light

aromatic tars (non-condensable tars) such as benzene, toluene, etc. (i.e., Group 1 in the Figure

3.16) will increase the heating value, and energy utilization can be improved. However,

condensable tars (light PAH tars) such as naphthalene, anthracene, pyrene, etc. (i.e., Groups

2~4) can give damages to the machinery. Therefore, it should be concentrated to remove the

condensable tars.

Non-condensable tar, benzene, is not intended to be adsorbed at the sludge char to improve

the energy efficiency. The sludge char from the carbonization-activation could adsorb small

amount of benzene, which was easy to be absorbed into micropore, but breakthrough point

was reached in a short time. Therefore, it could be passed without showing unwanted

adsorption. In addition, development of mesopore in the sludge char well adsorbed

condensable tar. Therefore, it could be effective for condensable tar reduction in producer gas

from the pyrolysis and/or gasification. Phuphuakrat et al. also confirmed the preferred

adsorption of condensable light tar and moisture for adsorbent with mesopore using a wood

chip experiment [81].

3.4 Summary

To utilize the sewage sludge from waste water treatment plant as energy and resource, the

pyrolysis, the steam gasification and the carbonization-activation on the dried sewage sludge

was conducted. Characteristics of each case were evaluated according to the products (i.e.,

sludge char, producer gas, and tar) in detail.

The pyrolysis on dried sludge showed the formation of gas, tar, and char during the primary

pyrolysis along with volatilization of organic component, and gas conversion from tar and

char during the secondary pyrolysis. The energy yield was 109 kJ.

The steam gasification displayed the higher amount of gas and tar compared with pyrolysis

along with reduced amount of sludge char. Due to the steam reforming, H2 and CO contents

were significantly increased, and the energy yield was also increased to 291 kJ. In case of tar,

the total amount of tar was increased, but the amount of light tar was reduced. Porosity of the

sludge char was improved to show 77.5467 m2/g of specific surface area.

The carbonization-activation showed the slight reduction of gas and tar formation compared

with the steam gasification. For the producer gas, it was mostly composed of H2 and CO like

the steam gasification, but the energy yield was reduced to 226 kJ, compared with the steam

gasification. Porosity of the sludge char was relatively improved to display 80.28 m2/g of

specific surface area and 6.229 nm of average pore diameter. In addition, the light tar

concentration was smaller than the pyrolysis and the steam gasification.

Due to the nitrogen content in sludge during the pyrolysis and/or gasification, nitrogen-

containing tar, such as benzonitrile and benzeneacetonitrile, was formed. These could induce

air pollution as a precursor to NOx via conversion into ammonia (NH3) and hydrogen cyanide

(HCN).

51

The development of mesopore in the sludge char well adsorbed condensable tar, while the

non-condensible tar passed. Therefore, it could be effective for condensable tar reduction in

the producer gas from the pyrolysis and/or gasification.

In conclusion, it was found to be effective for the formation of organic volatilization in the

primary pyrolysis (i.e., carbonization) and the steam activation in the secondary gasification

to achieve higher porosity sludge char and clean producer gas. That is, the carbonization-

activation was the best option among the three cases.

52

Chapter 4

Designing and design verification

of a plasma-catalyst reformer The development of a novel reformer is currently important not only to convert pyrolysis and

gasification gas including tar into sustainable low-pollution recycling energy, but to settle the

global warming environmental problem.

In this Chapter 4, a gliding arc plasma reformer (GAPR) was designed and verified its

performance by using a representative surrogate biogas (CH4+CO2) which can produce from

digesters in a waste water treatment plant. The developed GAPR was used for tar destruction

performance in Chapter 5.

Parametric screening study was conducted for the variables that affect biogas reforming of the

GAPR, and presented the optimal operating condition for hydrogen-rich gas production.

The developed GAPR had a quick starting characteristics and response time, had a high

conversion rate, and maintained optimal operating status for maximizing the gas property. In

addition, it is open to the application of various kinds of biogas gas reforming and tar

destruction for pyrolysis and gasification gases.

4.1 Literature review Through a thermal decomposition process, wastes like sewerage sludge, biomass, solid waste,

etc., can be used as an alternative source of energy [12]. The biogas from the anaerobic

digestion process in waste water treatment contains the major components CH4 and CO2 [100].

For the biogas gas, CH4 and CO2 play a role of global warming gases, having an adverse

influence on environment. But when using the biogas after converting it into hydrogen-rich

gas, it is possible to use the biogas as a more stable, energy-efficient, and environment-

friendly fuel [101-103]. The converted hydrogen-rich gas can be utilized in IGCC (Integrated

Gasification Combined Cycle), and IGFC (Integrated Gasification Fuel Cell), etc. [100].

Currently plasma technology has been applying to hydrogen production as a new method

[104]. In general, plasma is classified into two kinds: the thermal plasma called as the

equilibrium plasma and the non-thermal plasma called as the non-equilibrium plasma.

For the thermal plasma, the electrical power supplied to plasma discharge is high (higher than

1 kW). The neutral species and electrons have then the same temperature (around

5,000~10,000 K). The temperature in the plasma reactor and the energy consumption are thus

very high. And the cooling of the electrodes is generally useful to reduce their thermal erosion

[30, 105]. The use of this technology is therefore not relevant for an efficient production of

hydrogen in terms of energy consumption.

For the non-thermal plasma, the electrical power is very low (few hundreds watts). The

temperature of neutral species does not change, whereas the temperature of electrons is very

high (up to 5,000 K). In this case, the role of the plasma is not to provide energy to the system

but to generate radical and excited species allowing initiating and enhancing the chemical

53

reactions. The advantages of using non-thermal plasma are related to the lower temperature

that will result in lower energy consumption and lower electrode erosion.

In this two general types of plasma discharges, it is impossible to simultaneously keep a high

level of non-equilibrium, high electron temperature and high electron density, whereas most

prospective plasma chemical applications simultaneously require a high power for high

reactor productivity and a high degree of non-equilibrium to support selective chemical

processes. These parameters are somewhat achievable in the gliding arc plasma. The gliding

arc occurs when the plasma is generated between two or more diverging electrodes placed in a

fast gas flow. The gliding arc discharge has such strong points of easy response control, high

energy efficiency, and environment-friendliness which could be developed into a new

alternative technology [106, 107]. The gliding arc plasma is advantageous for its compactness

and quick starting and responding characteristics, and is used to convert diverse bio gas, fossil

fuels, etc. It also has high conversion efficiency in the optimal condition [34].

The gliding arc plasma discharge can be divided into the following three stages, as shown in

Figure 4.1: (A) breakdown, (B) equilibrium heating phase, and (C) non-equilibrium reaction

phase [104]. The initial breakdown (A) of the processed gas begins the cycle of the gliding

arc evolution. The high-voltage generator provides the necessary electric field to break down

the gas between the electrodes, and the discharge starts at the shortest distance between the

electrodes. The equilibrium stage (B) takes place after the formation of a stable plasma

channel. The gas flow convects the resulting small equilibrium plasma volume, and the length

of the arc column increases with the voltage. The non-equilibrium stage (C) begins when the

length of the gliding arc exceeds its critical value. Heat losses from the plasma column begin

to exceed the energy supplied by the source, and it is impossible to sustain the plasma in the

state of thermodynamic equilibrium. After the decay of the non-equilibrium discharge, there

is a new breakdown at the shortest distance between the electrodes, and the cycle is repeated.

Figure 4.1 Phases of gliding arc evolution; (A) Breakdown; (B) Equilibrium heating phase;

and (C) Non-equilibrium reaction phase

In this study, a noval gliding arc plasma reformer (GAPR) was designed and experimentally

verified. To show the catalytic effect in hydrogen-rich gas production, the GAPR was

combined with a catalyst reactor.

Parametric screening study was conducted for the variables that affect reforming in a

surrogate biogas in the GAPR. And the optimal operating conditions were shown for

hydrogen-rich gas production.

54



Figure 4.2 shows an experimental setup used for plasma reforming tests. The setup was

composed of a GAPR, a gas feeding line, a power supply, a control device, and a measuring

line.

The GAPR is combined with a catalyst reactor. This has three knife-shape electrodes located

at 120 degree in a quartz tube. The three electrodes are fixed opposite as a gap of 4 mm on a

ceramic ring part. The quartz tube is used for the outer shell of the plasma reformer for the

purpose of insulation and internal observation. In addition, the gas jet nozzle is installed with

a diameter of 3mm at the upper part of electrodes. The catalyst reactor was designed in a

triple co-axial tube to preheat catalysts evenly, filled by the Ni catalyst (Sűd-chemie, FCR-4,

Japan) manufactured using the impregnation method, with spherical γ-Al2O3 of 2mm diameter

as a supporter.

The gas feeding line supplies surrogate biogas into the plasma reformer, using the CH4 MFC

(LINETECH, M3030V, Korea) and CO2 MFC (BRONKHOST, F201AC-FAC-22-V,

Netherlands). In terms of steam, the water supplied from the water tank is supplied into the

steam generator, using the quantitative pump (KNF, STEPDOS03, Switzerland).

For the power supply equipment, the power supply (Unicorn Tech, UAP-15KIA, Korea) was

used to stabilize plasma discharge within the plasma reformer, and has maximum capacity of

15 kW (voltage: 15 kV; AC: 1A). As the measuring equipment, a high voltage probe

(Tektronix, P6015, USA) and a current probe (Tektronix, A6303, USA) were installed to

determine electricity characterization for the plasma reformer.

The control-monitoring device was used by the LabVIEW(National Instrument, LabVIEW

8.6, USA) to control the MFCs and the water pump. Also, it was used for monitoring the

changes of temperatures and other conditions automatically.

Figure 4.2 Schematic of the GAPR setup

55

The measuring-analysis line was composed of the temperature measurement and gas analysis.

Temperature was measured, using the K-type thermocouple with a data logger (KIMO,

KTT300, USA). In terms of gas analysis, H2, CO and hydrocarbon gases (CH4, C2H4, C2H6)

were sampled and analyzed at the same time, using a sampling line and two gas

chromatographs (SHIMAZU, GC-14B, Japan; VARIAN, CP-4900, Netherlands).

4.2.2 Experimental methods

Before injecting the surrogate biogas into the GAPR, the steam generator was heated at the

temperature of 250℃. The catalyst reactor was also heated up to the set temperature, using an

external burner. After all the part was thermally stable, experiments were conducted for each

parameter as shown Table 4.1.

The surrogate biogas of CH4 and CO2 mixture was injected with their flow amount controlled

by MFC (mass flow controller). And the steam was flown from the steam generator together

with the biogas gas, and injected into the GAPR in a state of mixed gas. The steam amount is

calculated by the water amount controlled by a pump whose micro-control is possible.

The reforming gas was sampled at the sampling port installed at the exit of the catalyst reactor,

and the sampling gas was analyzed continuously by gas chromatographs in dry-basis after

passing through the glass wool and cooler to remove soot and moisture in the reforming gas.

TCD was used as a detector. For measuring H2; CO and CH4; C2H4, C2H6, and CO2 Molecular

Sieve 5A (SHIMAZU); Molecular Sieve 5A (VARIAN); Porapak Q (VARIAN), are used

respectively, as analysis columns.

The instantaneous voltage, current, and dissipatedelectric power that were observed with a

digital oscilloscopeas shown Figure 4.3, showed almost the random feature of the history of

each gliding breakdown powered by a 3-phasealternative current power supply. The electric

power should be measured by calculating the root mean square voltage and the root mean

square current wave.

Figure 4.3 Applied voltage and current waveform

An opposite variation tendency can be observed from the arc voltage and the arc current

signals in one period. At the steady-state plasma condition after the breakdown point, the

56

voltage decreased to below the adjusted original voltage. On the other hand, the current value

increased and became higher than before the breakdown. This phenomenon was caused by the

arc production in the plasma, which typically occurred at low voltage and high current

conditions [108, 109].

The parametric screening studies were carried out according to the changes of the steam feed

rate (i.e., steam/carbon ratio), the catalyst bed temperature, the total gas feed rate, the input

electric power, and the biogas content. Table 4.1 shows the experimental range for various

parameters. Also, test was conducted at optimal operating condition (Table 4.2) for obtaining

the highest hydrogen concentration in the reforming gas for each variable.

Table 4.1 The experimental ranges for each parameter

Experi-

mental

variables

Steam feed

rate (Seam/

Carbon ratio)

Catalyst bed

temperature

(℃)

Total gas

feed rate

(L/min)

Input electric

power (kW)

Biogas

content

(CH4:CO2)

Range 1~5.5 506~777 8~2.4 0.525~0.76 6:4 ~ 4:6

4.2.3 Data analysis

▌H2 yield [110] and H2 selectivity [111]

The amounts of products are given in different ways: mol, mol percentage, yield or selectivity.

The definition of the two last magnitude amounts is more ambiguous and has been therefore

taken in all calculations as following:

100×][H

][H

injected atoms H ofamount Total

Hformed thein atoms H ofAmount =(%) yieldH

gas feed2

syngas222 (4.1)

where [H2]syngas is the hydrogen amount (L/min) within the reforming gas, and [H2]feed gas is

the maximum hydrogen amount (L/min) converted from the surrogate biogas:

product formed in the atoms H ofAmount

H formed in the atoms H ofAmount =(%)y selectivit H 2

2

100×O][H]2[CH

][H

converted2converted4

syngas2

(4.2)

where [H2]syngas is the hydrogen amount (mol) within the reforming gas, [CH4]converted is the

conversion amount (mol) of CH4 from the surrogate biogas and [H2O]converted is the conversion

amount (mol) of H2O from the steam.

Since one mol of methane can be converted into two mols of hydrogen as shown in Eq. 4.2,

100% hydrogen selectivity means that all hydrogen atoms of the methane molecules are

converted into hydrogen molecules, which can be obtained in ideal reforming processes.

▌CH4 conversion rate [31]

In order to produce hydrogen, the hydrocarbon molecule in the methane (CH4) has to be

cracked, to break the C-H links. The performance of this operation is evaluated by using the

CH4 conversion rate which is the ratio of CH4 contained in reforming products to the CH4

contained in the surrogate biogas:

57

100×][CH

][CH][CH=(%) MCR

input4

output4input4 (4.3)

where is the methane input amount (L/min), and is the methane

output amount (L/min).

▌Energy conversion efficiency [31]

The energy conversion efficiency of a GAPR-catalyst reactor is the sum of the lower heating

value (LHV) of hydrogen multiplied by the amount of hydrogen produced and that of carbon

monoxide multiplied by the amount of carbon monoxide produced divided by the input

energy, that is the summation of the electrical energy of the plasma discharge, the heating

energy of the steam generator and the LHV of the hydrocarbon injected multiplied by its

amount:

100×) LHV(CHFUEL HE IEP

LHV(CO)×[CO])LHV(H×][H= (%) ECE

4injectedenergyenergy

produced2produced2

(4.4)

where [H2]produced is the production amount of hydrogen (m3/h), LHV (H2) is the lower heating

value of hydrogen (kJ/Nm3), [CO]produced is the production amount of carbon monoxide (m

3/h),

LHV (CO) is the lower heating value of carbon monoxide (kJ/Nm3), IEPenergy is the plasma

input electric power (kJ/h), HEenergy is the heating energy of the steam generator (kJ/h),

FUELinjected is the methane amount in surrogate gas (m3/h) and LHV (CH4) is the lower

heating value of methane (kJ/Nm3).

It is expected that the entire CO produced is then converted into H2 by water-gas shift (WGS)

reaction (Eq. 4.10). Therefore, the CO produced can be taken into account for the calculation.

▌Specific energy requirement [31]

This value is the input electric power used by the plasma discharge that is required for

producing one mol of H2. Still considering the CO produced, the specific energy requirement

is expressed by the following Eq. 4.5.

produced2

energy

CO][H

IEP=)SER(kJ/mol

(4.5)

where energyIEP is the input electric power (W) fed to a GAPR, and [H2+CO]produced is the

production amount of the synthetic gas (L/min).


In this study, tests were conducted to clarify the optimal operating condition for maximizing

the hydrogen content in the reformed gas, by using the GAPR. The respective conditions and

test results for the optimal condition are shown in Table 4.2.

The results including the CH4 conversion rate of almost 100 % and the specific energy

requirement of 63 kJ/mol, proves that the GAPR developed in this study was designed well,

compared to other study having the results of the CH4 conversion rate of 95 % and the

specific energy requirement of 1,270 kJ/mol [112].

58

Table 4.2 Optimal conditions and their results

Conditions

Steam

/Carbon

ratio

Catalyst bed

temperature

(℃)

Total gas

feed rate

(L/min)

Input

electric

power

(kW)

Biogas

content

(CH4: CO2)

Value 3 700 16 0.525 6:4

Syngas concentrations

(dry vol.%)

CH4

conv.

rate(%)

H2

selectivity

(%)

H2 yield

(%)

Energy

conversion

efficiency

(%)

Specific energy

requirement

(kJ/mol)

H2 CO CO2 CH4 100 59 59 94.3 63

62 8 27 0

The CH4 conversion can be basically described by the plasma cracking, the steam reforming

and the dry reforming [102, 113].

▰ CH4 plasma cracking

CH4 ⇋ C +2H2 ΔH = +75 kJ/mol (4.6)

The generated carbon (C) can be converted to H2 and CO with CO2 and H2O acting as

gasifying agents.

C +CO2 ⇋ 2CO ΔH = +172 kJ/mol (4.7)

C +H2O ⇋ CO +H2 ΔH = +132 kJ/mol (4.8)

▰ Steam reforming reaction

CH4+H2O ⇋ CO +3H2 ΔH = +206 kJ/mol (4.9)

▰ Water gas shift reaction

CO +H2O ⇋ CO2+H2 ΔH =-41 kJ/mol (4.10)

▰ Dry reforming reaction

CH4+CO2 ⇋ 2CO+2H2 ΔH = +247 kJ/mol (4.11)

The parametric screening studies were carried out by changing the steam feed rate, the

catalyst bed temperature, the total gas feed rate, the input electric power, and the biogas

content. Each test was conducted by changing its value, while other variables were fixed at

the conditions shown in Table 4.2.

1) Effects of the steam feed rate

Figure 4.4 shows test results for the changes of the steam feed rate. The effect of the steam

feed is expressed as a steam to carbon ratio (i.e., S/C ratio). The other variables, the catalyst

bed temperature, the total gas feed rate, and the input electric power were set at 700℃, 16

L/min, and 2.4 kW respectively which is the optimal condition as shown in Table 4.2.

Through the pretest for the developed GAPR, carbon black (i.e., coke) was formed in the

reformer at the S/C ratio of 1 or lower, and the temperature of the steam generator was

reduced by the increase of the amount of injecting water at the S/C ratio of 5.5 or higher.

Therefore, tests were conducted with the S/C ratio set at 1~5.5.

59

Figure 4.4(a) shows the concentrations of H2 and CO in accordance with S/C ratio. Tests were

conducted with catalyst or without catalyst in the catalyst reactor.

For with catalyst in the catalyst reactor, the concentration of H2 was increased with the

increase of S/C ratio until reaching the optimal value of 62% at 3. The reason of increasing H2

at the region of low S/C ratio is that the steam reforming reaction of Eq. 4.9 was prevailed

over. Thereafter, H2 amount was almost kept constant. This is because the excessive injection

of steam has little influences on the production of hydrogen.

The concentration of CO was slightly decreased up to S/C ratio of 3 due to the water-gas shift

(WGS) reaction (Eq. 4.10). After this point, the CO maintained almost unchanged due to little

effectiveness of the excessive steam injection.

In addition, the catalyst improves the efficiency of reforming, with the optimum hydrogen

concentration increased from 40% to 62%, as compared to without catalyst. On the other hand,

the concentration of CO decreased from 19% to 8%.

Figure 4.4(b) shows the CH4 conversion rate, H2 yield and H2 selectivity by steam injection in

case of with catalyst in the catalyst reactor.

The CH4 conversion rate slightly increased with the increase of the S/C ratio, amounting to

almost 100% at the S/C ratio of 3. After this maximum value, this maintained almost the same

values, showing that the fed CH4 gas was mostly converted into the synthetic gas. Both the H2

yield and the H2 selectivity had similar pattern with the H2 concentration.

(a) Concentrations of reforming products (b) CH4conversion rate, H2yield, and H2

selectivity

Figure 4.4 The effect of the steam feed rate

2) Effects of catalyst bed temperature Figure 4.5 presents the reforming characteristics by changing the catalyst bed temperatures of

506~777℃. The other variables like the S/C ratio, the total gas feed rate, and the input electric

power were fixed at 3, 16 L/min, and 2.4 kW, respectively.

Figure 4.5(a) shows the concentrations of selected reforming products with the variation of

the catalyst bed temperature. The H2 concentration significantly increased with the increase of

the catalyst bed temperature, showing the maximum value of 62% at 700℃. But at the

temperature above 700℃, the concentration was slightly decreased. At the high temperature,

Steam/Carbon ratio

H2

co

ncen

tratio

n(%

)

CO

co

ncen

tratio

n(%

)

1 2 3 4 5 60

20

40

60

80

0

20

40

60

80

100With catalyst

Without catalyst

H2

H2

CO

CO

(a)

CH

4co

nvers

ion

rate

(%)

50

60

70

80

90

100

Steam/Carbon ratio

H2

sele

ctivity

(%)

H2

yie

ld(%

)

1 2 3 4 5 620

40

60

80

100

0

20

40

60

80

CH4

conversion rate

H2

yield

H2

selectivity

(b)

60

the increase of H2 production from the steam reforming (Eq. 4.9) and the dry reforming (Eq.

4.11) reactions is lower than the decrease of H2 production from the water gas shift reaction

(Eq. 4.10; equilibrium is shifted towards CO and H2O). This leads to a decrease of the H2

content, while CO increased and CO2 decreased. The CH4 concentration was 9% at the initial

catalyst bed temperature of 506℃. The concentration decreased gradually and then it was 0%

at the temperature of 700℃, which should be the optimal condition.

Figure 4.5(b) shows the CH4 conversion rate, the H2 yield and the H2 selectivity by the

variation of the catalyst bed temperature. Increasing the catalyst bed temperature implies a

slight increase of the CH4 conversion until 700℃ is reached. Above 700℃, the gain in

conversion is very low, showing almost 100%. Therefore, the catalyst bed temperature should

not be below 700℃ because from there the CH4 conversion decreases. The H2 yield and the

H2 selectivity increased by 59% and 31% respectively at 700℃ which was the optimal

operating condition.

(a) Concentrations of reforming products (b) CH4 conversion rate, H2 yield and H2

Selectivity

Figure 4.5 The effect of the catalyst bed temperature

3) Effects of the total gas feed rate

Figure 4.6 presents the test results in effects of the total gas feed rate. Gas was injected with

the total gas feed rates of 8~24 L/min after fixing the S/C ratio, the catalyst bed temperature,

and the input electric power at 3, 700℃, and 2.4 kW, respectively.

Figure 4.6(a) shows that the increase of the total gas feed rate causes the H2 concentration to

decrease from 63% to 55%, but causes the CO concentration to increase from 2% to 17%.

CH4 and CO2 increased with increasing the total gas feed rate. The increase in the total gas

feed rate led to the decrease of the residence time in the GAPR and the catalyst reactor.

Therefore, CH4 and CO2 were increased with lower steam reforming and dry reforming

reactions expressed by Eqs. 4.9 and 4.11.

Figure 4.6(b) shows the CH4 conversion rate, the H2 yield, and the H2 selectivity. The increase

of the total gas feed rate caused the CH4 conversion rate to decrease from 100% to 99%.The

increase of the total gas feed rate caused the H2 yield to decrease from 63% to 55%, and the

H2 selectivity to decrease from 32% to 29%.

Catalyst bed temperature (oC)

H2,C

O,C

O2

co

ncen

tratio

ns

(%)

CH

4co

ncen

tratio

n(%

)

480 520 560 600 640 680 720 760 8000

10

20

30

40

50

60

70

0

4

8

12

16

20

CO

CH4

H2

CO2

(a)

Catalyst bed temperature (oC)

H2

sele

ctivity

(%)

H2

yie

ld(%

)

480 520 560 600 640 680 720 760 80020

30

40

50

60

70

80

90

100

0

10

20

30

40

50

60

70

80CH

4conversion rate

H2

selectivity

H2

yield

(b)

CH

4co

nvers

ion

rate

(%)

0

10

20

30

40

50

60

70

80

90

100

61

(a) Concentrations of reforming products (b) CH4 conversion rate, H2 yield, and H2

Selectivity

Figure 4.6 The effect of the total gas feed rate

4) Effects of the input electric power Figure 4.7 shows the influences by the changes of the input electric power. The changes of the

input electric power were made after fixing the S/C ratio, the catalyst bed temperature, and the

total gas feed rate at 3, 700℃, and 16 L/min respectively. Tests were conducted, with the

input electric power set at 2.4~3.5 kW.

Figure 4.7(a) presents selected gas concentrations, the CH4 conversion rate, the H2 yield, and

the H2 selectivity. The increase of the input electric power kept H2 and CO concentrations

almost constant at about 62% and 8%, respectively. The CH4 conversion rate was almost

100%, showing that methane is mostly converted into syngas. The H2 yield and the H2

selectivity were 59% and 31%, respectively, showing no great change.

Figure 4.7(b) shows the energy efficiency and the specific energy requirement caused by the

variation of the input electric power. The energy efficiency was almost the same value in

about 52% with increasing the input electric power. But the increase of the input electric

power caused the specific energy requirement to increase from 289 kJ/mol to 297 kJ/mol.

As a result, when taking the CH4 conversion rate and the energy efficiency into consideration,

it seems appropriate to apply the minimum input electric power of 2.4 kW as the optimal

operation condition in this study.

5) Effects of the biogas content Figure 4.8 shows the influences caused by the changes of the biogas content. The S/C ratio,

the catalyst bed temperature, the total gas feed rate and the input electric power were kept at 3,

700℃, 16 L/min and 2.4 kW, respectively. Tests were conducted with the biogas content set

at 6:4~4:6 based on the ratios of CH4 and CO2.

Total gas feed rate (L/min)

H2,C

O,C

O2

co

ncen

tratio

ns

(%)

CH

4co

ncen

tratio

n(%

)

8 12 16 20 240

10

20

30

40

50

60

70

80

0

0.5

1

1.5

2

CO

H2

CO2

CH4

(a)

CH

4co

nvers

ion

rate

(%)

90

92

94

96

98

100(b)


H2

sele

ctivity

(%)

H2

yie

ld(%

)

8 12 16 20 2430

40

50

60

70

80

90

100

0

20

40

60

80

100

CH4

conversion rate

H2

selectivity

H2

yield

62

(a) Selected gas concentrations,H2 yield, (b) Energy efficiency and specific energy

CH4 conversion rate and H2 selectivity requirement

Figure 4.7 The effect of the input electric power

Figure 4.8(a) shows the selected gas concentration in the reforming gas. When the CO2 ratio

in the biogas increased, the H2 concentration decreased from 62% to 48%. But the CO2

concentration increased from 25% to 46%. As we can expect, this is due to lower CH4 and

higher CO2 than the optimal condition (Table 4.2).

Figure 4.8(b) shows the CH4 conversion rate, the H2 yield, and the H2 selectivity. With the

increase of the CO2 ratio, the CH4 conversion rate was approximately 100%, showing that the

conversion was almost perfectly made. The H2 yield was slightly decreased from 58% to 51%,

and the H2 selectivity was also decreased from 31% to 24%.

As a result, the H2 yield, and the H2 selectivity were influenced by the change of the biogas

content, but the CH4 conversion rate showed no great change.

(a) Concentrations of reforming products (b) CH4 conversion rate, H2 yield, and H2

Selectivity

Figure 4.8 The effect of the biogas content

Input electric power (kW)

H2,C

Oco

ncen

tratio

ns

(%)

2.2 2.4 2.6 2.8 3 3.2 3.4 3.60

20

40

60

80

100

CH4

conversion rate

H2

yield

CO

H2

selectivity

H2

H2yie

ld,C

H4

co

nvers

ion

rate

(%)

H2sele

ctivity

(%)

0

20

40

60

80

100

30

40

50

60

70

80

90

100

(a)

Input electric power (kW)

En

erg

yeff

icie

ncy

(%)

Sp

ecific

en

erg

yre

qu

irem

en

t(k

J/m

ol)

2.2 2.4 2.6 2.8 3 3.2 3.4 3.630

40

50

60

70

80

90

100

0

80

160

240

320

400(b)

CH4

: CO2

H2,C

O2,C

Oco

ncen

tratio

ns

(%)

0

20

40

60

80

H2

CO2

CO

6 : 4 5.5 : 4.5 5 : 5 4.5 : 5.5 4 : 6

(a)

CH4

: CO2

0

20

40

60

80

100

H2yield (%)

CH4conversion rate (%)

H2selectivity (%)

6 : 4 5.5 : 4.5 5 : 5 4.5 : 5.5 4 : 6

(b)

63

In summary, the plasma technology is generally known to require a lot of energy. But the

energy consumption depends on the types of plasma discharge to be used. In addition, the

important factor for achieving high energy efficiency will be how to operate or combine other

auxiliary rigs.

Table 4.3 represents a comparison of this study other researches. This study shows higher

energy conversion efficiency than other studies, showing that the plasma-catalyst reformer (of

this study) can apply without any problem in view of the energy requirement, showing the

quick starting characteristics and the response time.

By the way, for the comparison of this plasma reformer with other systems such as using

thermal decomposition, the electric power should be converted to the primary energy.

Otherwise the comparison will be unfair, because the electric reformer have already

consumed much more energy to get the electric power. In Japan, the correct factor of 36.9% is

often used to convert the primary energy to the electric energy.

Table 4.3 Comparison of this study with other researches for fuel reforming [31, 114]

Researcher Bromberg et

al. Ahmar et al. Fidman et al.

Czernichowski

et al.

Thomas

Hammer et al. This study

Plasma

discharge type

AC Plasmatron

Gen3

Sliding Arc

Plasma

DC Gliding

Arc Plasma

DC Gliding

Arc Plasma

Dielectric

Packed Bed

Reactor

3 phase AC

Gliding Arc

Plasma with

Catalyst Bed

Oxidizing agent Air Air & Steam Air Air Steam Steam

Feedstock Methane Methane Methane Methane Methane Biogas

(CH4+CO2)

Input electric

power (kW) 0.375 0.83 0.05 0.36 0.6 0.525

Fuel conversion

efficiency (%) 73.38 22.53 87 100 68 100 (86.5)

Energy

conversion

efficiency (%)

36.7 24.59 75.81 50 60 94.3 (65.2)

<Note> ( ) is the result for plasma discharge only (i.e., without catalyst bed)

Figure 4.9 shows a comparison of the energy conversion efficiency between each research.

The figure contains information on the hydrocarbon feedstock, the non-thermal plasma device

as well as the institution involved in the development of the plasma technology. Efficiency

distribution is widely spread from 0.49% to 79%. The highest values correspond mainly to arc

discharge. The GAT (Gliding Arc in Tornado) reactor achieves the top value with 79%.

Compared to the researches, this study showed high level in the efficiency. Particularly, the

use of the catalysis bed gave highest efficiency as 94.3%.

64

Figure 4.9 Energy conversion efficiency studied by other researchers [31]

4.4 Summary

A gliding arc plasma reformer (GAPR) was designed and verified for the performance in

producing hydrogen-rich gas to reform the biogas including CH4 and CO2.

The optimal operating conditions and their results showed the concentrations of 62% H2, 8%

CO, 27% CO2, and 0% CH4 on the basis of the steam/carbon ratio of 3, the catalyst bed

temperature of 700℃, the total gas feed rate of 16 L/min, the input electric power of 2.4 kW

and the biogas content of 6:4. Also, the CH4 conversion rate was almost 100%, and the H2

yield and the H2 selectivity were 59% and 59%, respectively. At this time, the energy

conversion efficiency was 94.3 %, and the specific energy requirement was 63 kJ/mol.

To verify the performance of the GAPR, the parametric screening studies were conducted. In

the catalyst reactor, the concentration of H2 was increased with the increase of the S/C ratio

until reaching the maximum value of 62% at 3. Thereafter, the H2 was almost maintained

constant. The catalyst in the reactor improved the efficiency of reforming to 22% compared to

without the catalyst bed. The optimum production of H2 and the energy efficiency were

achieved at the catalyst bed temperature of 700℃. Therefore, the catalyst bed temperature

should not be below 700℃ because from there the CH4 conversion decreased. The increase

of the gas feed rate caused the reduction of the residence time in the GAPR and the catalyst

reactor, decreasing the H2 concentration, the H2 yield and the CH4 conversion rate. The

increase of the input electric power kept H2 and CO concentrations almost constant. In terms

of the biogas content, the H2 concentration decreased with the increase of the CO2 amount in

the biogas.

65


conversion rate, and maintained optimal operating status for maximizing the gas property. In

addition, it is open to the application of various kinds of light gas reforming and tar

destruction in pyrolysis and/or gasification gases.

66

Chapter 5

Plasma reformer performance for tar destruction The pyrolysis and gasification is an energy conversion technology for diverse waste resources,

including sewage sludge, biomass, and urban solid waste to produce synthetic gases for

industrial use. The tar in the thermal decomposition gas from the pyrolysis or gasification

process, however, damages synthetic gas facilities and causes operation trouble.

A gliding arc plasma reformer (GAPR) for tar decomposition was developed to address the

aforementioned problem. Benzene and anthracene were selected as a light aromatic tar and a

light PAH tar, respectively. Experiments were performed on the parameters that affect the tar

decomposition efficiency, and the optimal operation condition was presented.

To verify the performance of the GAPR for real tar, a continuous-type screw pyrolyzer was

designed and used for tar removal test at the optimal condition. Tar was sampled and analyzed

for the gravimetric tar and wet group light tars.

In addition, an externally oscillated plasma reformer (EOPR) was designed to enhance the

idea of the plasma reformer. Its performance for the tar destruction was achieved for light

aromatic tar (i.e., benzene). To identify the characteristics of the influential parameters of tar

decomposition, tests were performed on the oscillation frequency, the oscillation amplitude,

the steam feed rate, and the total gas feed rate.


Through a thermal decomposition gasification process, biomass, solid waste, organic

sewerage sludge etc can be used as an alternative energy source [12]. The producer gas

formed from biomass gasification contains the major components CO, H2, CO2, CH4, and N2,

in addition to organic (tars) and inorganic (H2S, HCl, NH3, alkali metals) impurities and

particulates.

The organic impurities range from low molecular weight hydrocarbons to high molecular

weight hydrocarbons. The lower molecular weight hydrocarbons can be used as a fuel in gas

turbine or engine applications. The higher molecular weight hydrocarbons are collectively

known as tar. However, tar formation during the thermal pyrolysis or the gasification is a

major problem for adoption. At ambient conditions, tar condenses or polymerizes into more

complex structures in exit pipes, heat exchangers and particulate filters, leading to choke and

attrition, which can result in the decrease of the total efficiency and an increase in the cost of

the process. Therefore, the aspect of tar cracking or removal during gas cleaning-up is one of

the most important technical uncertainties in implementation of the gasification technology

[115].

To solve this issue, numerous researches have been conducted, and they are divided into

physical and chemical methods. As a physical approach, scrubber, cyclone filter, wet-type

electrical dust collector, activated carbon, etc are available. Regarding chemical methods,

catalysis, thermal cracking, partial oxidation, plasma discharge, etc. can be the best possible

solutions. Scrubber is suitable for large-sized plant and for low environmental contamination,

67

and its cost is one of the most significant matter. Catalysis method is applicable but at risk to

sulfuric, chlorine, and nitrogen compounds, and coke is easy to form.

So, recent researchers using the plasma discharge have been taking initiative to overcome

these problems. Especially, non-thermal plasma technology should be used to destruct the tar

under low pressure or atmospheric state with low power consumption [116].

After fundamental studies on the pulsed non-thermal plasma cracking for tar removal, higher

efficiency of tar removal has been exhibited due to the formation of radical in comparison to

the existing thermal and catalytic cracking [117]. However, the installation cost and short life

cycle of the pulse power supply is the key for implementation. Besides these studies removal

of VOCs (volatile organic compounds) like light aromatic tars, such as benzene, toluene,

xylene, etc are required to be investigated. Reduction technologies of VOCs using plasma

technologies are mainly based on the corona discharged, the dielectric barrier discharged

(DBD), the gliding arc discharged, etc. These methods show a high energy efficiency, and are

not affected by the type and concentration of VOCs. This feature gives additional attention to

public [118-120].

However, the corona discharge and DBD has significant effects on the reactor flow rate, and

the density of plasma is relatively low. They can be applied to scientific research, but

commercial potential is low. In addition, selectivity during the reforming reaction is difficult

to control. Meanwhile, the gliding arc discharge features quick start-up performance within

few seconds, and easier to control reactions. Along with these, higher destruction efficiency

and lower energy utilization can be achieved. This method is developing as a new energy

alternative [104, 121].

A GAPR which was developed in the biogas reforming study (Chapter 4) was used for tar

destruction. Parametric studies on the factors that can affect the decomposition and

destruction energy efficiency of benzene and anthracene as representative tar were conducted

to know tar destruction characteristics. Through the parametric study, the optimal conditions

and their results were taken to guide operating patterns. And the GAPR was verified for

destructing real tar produced in a continuous pyrolyzer.

In addition, an externally oscillated plasma reformer (EOPR) was designed to enhance the

idea of the plasma reformer. Its performance for tar destruction was achieved for light

aromatic tar (i.e., benzene).



Figure 5.1 shows the test equipment diagram for the tar removal test rig. The equipment

consisted of a GAPR, a steam feeding line, a tar feeding line, a power supply equipment, a

measuring-analysis line, and a control-monitoring system.

The GAPR was made of a 55 mm-diameter and 200 mm-long quartz tube so that the

insulation could be ensured and the interior could be checked. Three knife-shaped electrodes

were fixed around the center of the GAPR at 120 degree by the ceramic support (Al2O3 ; 96

wt.%). An injection nozzle for tar-gas mixture was installed at the center of the upper part of

the support. The reformer was designed so that diverse internal parts could be replaced

(electrode lengths: 70 mm, 95 mm, and 125 mm; electrode gaps: 2 mm, 3 mm, and 4 mm; gas

68

nozzle diameters: 1.5 mm, 3 mm, 4 mm, and 5 mm; and electrode shapes: triangle, Arc 1, Arc

2, and Arc 3).

The steam feeding line consisted of a steam generator and a water pump (STEPDOS 03, KNF,

Switzerland). The distilled water in the water tank was controlled by the water pump and fed

into the steam generator. After transformed into steam, it was fed into the GAPR together

with the dilution gas (nitrogen).

The tar feeding line had a tar generator that consisted of a mantle heater, a tar container, and

an MFC (M3030V, LINETECH, Korea) that controlled the feed rate of the surrogate tar

carrier gas, which was nitrogen. The generated tar in the container was fed into the GAPR

while the mantle heater temperature and the carrier gas feed rate were controlled.

The power supply equipment consisted of a power supply (UAP-15K1A model, Unicon Tech.,

Korea), a high-voltage probe (P6015, Tektronix, USA), a low-current probe (A6303,

Tektronix, USA), and an oscilloscope (TDS-3052, Tektronix, USA) for electrical

characteristic measurement. The power supply provided power of up to 15 kW (voltage: 15

kV and AC current: 1 A) as a three-phase AC (alternative current) to the GAPR. The power

was measured using the voltage and current probes.

The measurement-analysis line consisted of sampling parts and analysis equipments. The

sampling parts consisted of a soot filter (LS-25, Advantec, Japan), impingers, a flow meter

(RMA-10, Dwyer, U.S.A), a gas meter (W-MK-10-ST, Shinagawa, Japan), and a suction

pump (N-820.3FT 18, KNF, Switzerland). The analysis equipments used were a GC-FID

(GC-14B, SHIMADZU, Japan) for tar analysis and a GC-TCD (CP-4900, Varian,

Netherlands) for gas analysis. The cotton and active carbon filters was installed to protect the

GC-TCD column from the remaining tar and VOCs.

The control-monitoring device was connected to relevant parts, including the MFC, water

pump, heater, and soot filter, and those parts were controlled using LabVIEW (National

Instrument LabVIEW 8.6, USA) on a computer. This system enabled continuous monitoring

of the temperature, steam flow rate, and nitrogen gas feed rate.

Figure 5.1 Schematic diagram of the experiment setup

69

5.2.2 Experimental methods

Figure 5.2 shows the initial operating characteristics and stabilization conditions showing

temperatures at the optimal condition for benzene tar at each part. The temperature of the

steam generator① was kept constant at 300℃. The temperature of the tar generator② was

kept at 25℃ because the boiling point of benzene is 80.1℃ and it can vaporize at the room

temperature. The heating line③ was heated up to 100℃ using a tape heater to prevent

condensation. The temperature of the GAPR④ was maintained at about 290℃. The

temperature of the soot filter⑤ was kept constant at 120℃.

For the anthracene test, the final stable temperatures at each part were set as follows; the

steam generator 490℃, the tar generator was at 260℃, the heating line was at 400℃ due to

boiling point of anthracene (=340℃), the GAPR 375℃ and the soot filter was at 120℃.

The tar carrier gas was fed into the tar generator that contained liquid surrogate tar (benzene

or anthracene) at the fixed temperature. The surrogate tar was vaporized, and a stable tar-

containing gas was generated. Water and the dilution gas were fed into the steam generator

and heated to the set temperature to generate steam. The generated steam and tar-containing

gas were mixed in the orifice mixer, and the mixture was fed into the GAPR. The test was

continued under the stable plasma discharge condition, maintaining a constant temperature at

each component.

Figure 5.2 Initial operating characteristics and stabilization conditions for benzene tar at each

component

The tested operating parameters that affect the tar decomposition and the destruction energy

efficiency were: the steam feed rate, the input benzene concentration, the total gas feed rate,

and the specific energy input (SEI). Table 5.1 and Table 5.2 show the ranges of the parameter

values for the benzene and anthracene, respectively. In addition, to know the design factors, in

Time(min)

Te

mp

era

ture

(oC

)

0 15 30 45 60 75 900

50

100

150

200

250

300

350

1 Steam ganearator

3 Heating line

4 GAPR

2 Tar generator

5 Soot filter

Starting point for experiment

Starting point for plasma discharge

70

the benzene test the nozzle diameter, the electrode gap, the electrode length, and the electrode

shape were changed as shown in Table 5.1. Figure 5.3 shows the detailed electrode shapes.

Table 5.1 Experiment conditions for benzene tar

Conditions Steam feed

rate (L/min)

Input

benzene

conc. (%)

Total gas

feed rate

(L/min)

Specific energy

input (kWh/m3)

Nozzle

diameter

(mm)

Electrode

gap (mm)

Electrode

length

(mm)

Electrode

shape

Variables

range 0~0.85 0.07~0.25 12.6~24.7 0.17~0.36 1.5~5 2~4 70~125

Triangle,

Arcs 1~3

Table 5.2 Experiment conditions for anthracene tar

Conditions Steam feed rate (L/min) Input anthracene conc.

(g/Nm3)

Total gas feed rate

(L/min)

Specific energy

input (kWh/m3)

Variables range 0~1.57 0.1~0.68 7.2~30.1 0.175~0.234

Figure 5.3 Types of electrode shapes

To measure the input surrogate tar, the input gas mixture was sampled from the inlet of the

GAPR. And the surrogate tar, the carbon-black and the reforming gas were sampled from the

outlet of the GAPR.

Tar sampling and analysis were conducted for benzene by the Industrial Standard [122] and

for anthracene by the wet-type Biomass Technology Groups (BTGs) [82].

For the benzene tar, the soot and moisture contents in the sampling gas were removed by

glass wool and calcium chloride (CaCl2), and the sampling gas was collected from the

benzene sampling port using a syringe (22265, Supelco, USA). The collected benzene tar was

injected into a flame ionization detector (FID) port of the gas chromatograph using a RTX-5

(RESTEK) capillary column (30 m-0.53 mm id, 0.5 μm film thickness) for the analysis. The

FID analysis conditions were as follows: the temperatures of the injector and the detector

were kept constant at 200℃ and 280℃, respectively. The oven temperature increased at a rate

of 10℃/min within the range of 40~80℃ and 20℃/min within the range of 80~300℃, and

then the oven was left for 5 min for stabilization.

For the anthracene tar, the wet-type tar sampling and analysis [82] were conducted. Three

impingers were separately installed in two baths. The temperature of the first water bath was

71

kept constant at 20℃ or below, and two impingers filled with 50 mL of isopropanol were

installed. The temperature of the second isopropanol bath was kept constant at -20℃ or below

using a chiller, and an empty impinger was installed. The tar in the gas was condensed and

collected in the impingers in the two baths. A suction pump (N-820.3FT 18 model, KNF,

Switzerland) was used for the collection of tar and steam at the flow rate of 3 L/min for 20

min.

The tar solution that was collected in the impinger was analyzed using the GC-FID. An RTX-

5 column (RESTEK, USA; 30 m-0.53 mm id; 0.5 μm film thickness) was used for the tar

analysis. The oven temperature was kept constant at 45℃ for 2 minutes, and was increased at

the rate of 7℃/min to 320℃ and then was maintained for 2 minutes. The temperatures of the

detector and the injector were set at 340℃ and 250℃, respectively.

To measure the carbon-black concentration, soot was collected from the glass filter paper

(GA-100, Advantec, Japan) on the soot filter. The difference between the sampled soot

weights before and after the sampling was measured using an electronic balance (HS-250D,

Shenyang Longteng, Taiwan). To know the accumulated gas amount, a gas meter was used

for 20 minutes at a sampling gas flow rate of 2.5 L/min.

To analyze the reforming gas, the GC-TCD (CP-4900, Varian, Netherlands) was used with

the Molecular sieve 5A column (for H2, CO, O2, and N2) and the PoraPlot Q column (for CO2,

C2H4, and C2H6).

5.2.3 Data analysis

▌Decomposition efficiency

The decomposition efficiency, which represents the degree of tar destruction in the producer

gas, was calculated using Eq. 5.1, as follows:

100[VC]

[VC][VC](%)η

inlet

outletinlett

(5.1)

where inlet[VC] is the input tar concentration (%) and

outlet[VC] is the output tar concentration

(%).

▌Destruction energy efficiency

The destruction energy efficiency was calculated using Eq. 5.2 [123].

IP

Q )[MC]([MC](g/kWh)η outletinlet

e

(5.2)

where inlet[MC] is the input tar concentration (g/m

3) and

outlet[MC] is the output tar

concentration (g/m3). Q is the gas feed rate (m

3/h) for the reformer and IP is the input electric

power (kW).

▌Specific energy input The specific energy input (SEI), which is the ratio of the input electric power to the gas feed

rate, was calculated using Eq. 5.3.

Q

IP)SEI(kWhm3 (5.3)

72

▌Carbon balance The carbon balance, which represents the carbon mass conservation, was calculated using Eq.

5.4 as follows:

100)[MC]A([MC]

ST]H2[C]H2[C][CH][CO[CO](%) CB

outletinlet

624242

(5.4)

where [CO] , ][CO2, ][CH4

, ]H[C 42, and ]H[C 62

are concentrations of each producer gas

(g/m3), ST is the carbon-black concentration (g/m

3), and A is the carbon constant, which is 6

for benzene and 14 for anthracene.

5.2.4 Reaction mechanism for tar destruction

The tar destruction mechanism in the plasma arc discharge can be explained by the following

reactions. The main reactions are the tar cracking (Eq. 5.5) and the carbon formation (Eq. 5.6)

[26], as follows:

▰ Tar cracking

pCnHx → qCmHy + rH2 (5.5)

▰ Carbon formation

CnHx → nC + (x/2)H2 (5.6)

where CnHx represents tar, such as the large molecular compounds, and CmHy represents

hydrocarbon with carbon number smaller than that of CnHx.

Eqs. 5.7~5.11 show the mechanisms of the production, utilization, and termination of the

radical and soot decomposition when the steam is fed into the plasma discharge [117, 124].

▰ Radical production

H2O → H++OH

*+e

- (5.7)

▰ Radical utilization

OH* + TAR → Products (5.8)

▰ Radical termination

OH* + CO → CO2 + H (5.9)

▰ Soot decomposition

Cx + OH* → Cx-1 + CO + 1/2H2 (5.10)

Cx + 2OH* → Cx-1 + CO2 + H2 (5.11)

The reaction of the gas generated after the tar destruction and the steam reforming is

explained by the water-gas shift reaction (Eq. 5.12) and the steam reforming reaction (Eq.

5.13) [125, 126], as follows:

▰ Water-gas shift reaction

CO + H2O → CO2 + H2 (5.12)

▰ Steam reforming reaction

CnHm + nH2O = nCO + (n + m/2) H2 (5.13)

where CnHm is the light hydrocarbon.

73


5.3.1 Effects of light aromatic and PAH tars

1) Destruction for light aromatic tar

A GAPR was developed to destruct the tar generated from the pyrolysis and/or gasification.

Test was performed at the optimal conditions for benzene decomposition and the destruction

energy efficiency of the developed GAPR was investigated to verify the benzene destruction

performance. Table 5.3 shows the conditions and the experimental results.

The benzene decomposition efficiency was 82.6%, and the destruction energy efficiency was

20.9 g/kWh. H2, CO, and CO2 were mostly generated as the reforming gases converted from

the benzene light aromatic tar. The light hydrocarbons (CH4, C2H4, and C2H6) were converted

to CO and H2 due to the steam reforming (Eq. 5.13). Carbon-black was not generated because

its formation was suppressed by the soot decomposition under the sufficient steam condition

(Eqs. 5.10 and 5.11).

The carbon balance was calculated to be 91.4% according to Eq. 5.4. In this case, the value

was far from 100%, even though the line adsorption and analysis errors were considered. This

seems to be because of the following two reasons. First, the heavy hydrocarbon at the outlet,

which was analyzed with the GC, showed one micro-peak before the benzene peak. This was

an unknown material that consisted of carbon and hydrogen, which are lighter than benzene,

and it was not considered during the carbon balance calculation. Second, a part of the carbon-

black was converted into HCN and CN in the GAPR according to Eqs. 5.14 and 5.15. Tar

with a benzene ring cleavage or some intermediate products formed CH radicals, which in

turn formed HCN and CN radicals [123].

CH + N2 → HCN + N (5.14)

CH + N → CN + H (5.15)

Table 5.3 Test results for the optimal conditions

Optimal Conditions

Steam feed

rate (L/min)

Input

benzene

conc. (%)

Total gas feed

rate (L/min)

SEI1)

(kWh/m3)

Nozzle

diameter

(mm)

Electrode

gap

(mm)

Electrode

length

(mm)

Electrode

shape

0.66 0.12 16.7 0.17 3 3 95 Arc 1

Experiment Results

Result

Reforming gas composition (%)2) Carbon

black

(g/Nm3)

Carbon

balance

(%)

Higher

heating value

(kJ/Nm3)

Decomposition

efficiency (%)

Destruction

energy efficiency

(g/kWh) H2 CO CO2 CH4 C2H4 C2H6

38.9 33.4 27.6 0 0 0 0 91.4 9,209 82.6 20.9

Note: 1) Specific energy input; 2) Gas composition was excluded N2

In addition, this study was performed by changing the parameters which affect the benzene tar

destruction characteristics of the GAPR, The tests were performed according to the parameter

ranges shown in Table 5.1. The other parameters were fixed at their optimal values as shown

in Table 5.3.

74

▌ Effects of the Steam Feed Rate

The steam feed rate was changed from 0 L/min to 0.85 L/min with fixing other vatiables in

the optimal conditions (Table 5.3).

Figure 5.4(a) shows the benzene decomposition efficiency and the destruction energy

efficiency as a function of to the steam feed rate.

The decomposition efficiency was almost constant at 63% until the steam feed rate reached to

0.19 L/min. This was because such a small quantity of steam had almost no effect on the

decomposition efficiency. Without benzene was decomposed according to the tar cracking

(Eq. 5.5).

With the increase in the steam feed, the benzene decomposition efficiency gradually increased

and reached to 82.6% at the steam feed rate of 0.66 L/min. Then it decreased gradually. As

the steam feed increases in the plasma discharge, a large quantity of OH radicals is generated

by electrons by the strong energy, by the radical production reaction (Eq. 5.7). This is because

the generated OH radicals are converted into another material through the oxidative

decomposition of tar benzene, as in the radical utilization reaction (Eq. 5.8) [117]. The

dissociation energy (i.e., 9.8 eV) of the triple nitrogen bonding, N ≡ N, needs very energetic

electronic collisions, and the single water bonding, H–OH (i.e., 5.11 eV), is in the order of the

magnitude of the non-equilibrium electronic collisions in the plasma phase. Thus, dissociation

reactions of water molecules were accelerated in such higher-humidity gas plasma [127].

Water, however, also has an adverse effect on benzene decomposition due to its

electronegative characteristics. Too many water molecules limit the electron density in the

system and quench the activated chemical species [120]. Therefore, the benzene

decomposition efficiency reached its maximum and then decreased due to the increase in the

steam feed rate. In addition, with the increase in the steam feed, the overall gas in the GAPR

increased. Accordingly, a sufficient retention time was not ensured, and the benzene

decomposition efficiency decreased. The higher steam feed rate led to a stronger effect.

The destruction energy efficiency had a pattern that was similar to that of the benzene

decomposition efficiency. With the increase in the steam feed, the destruction energy

efficiency increased and reached to 20.9 g/kWh at the steam feed rate of 0.66 L/min. Then it

started to decrease. As shown in Eq. 5.2, the destruction energy efficiency increased due to

the increase in the tar destruction and the input feed rate with the increase in the steam for a

specific quantity of the input electric power. After the maximum destruction energy efficiency

was reached, it decreased because of the larger effectiveness of the tar destruction, despite the

increase in the gas feed rate.

Figure 5.4(b) shows the concentrations of carbon-black and light gases.

The quantity of carbon-black was relatively large (0.096 g/Nm3) when no steam was fed. This

was because the benzene tar was decomposed into carbon (C) due to the carbon formation (Eq.

5.6) without the oxidation caused by steam or oxygen. As the steam feed increased, carbon-

black gradually decreased and was hardly generated when the steam feed rate was 0.47 L/min

or higher. As in the soot decomposition reactions (Eqs. 5.10 and 5.11), this was because the

OH radical (Eq. 5.7) was converted into light gases (CO, CO2, and H2) [120].

The light gases were produced when the benzene tar was decomposed. With the increase in

the steam feed rate, H2 and CO2 continued to increase, and CO increased to its maximum and

then decreased. H2, CO2, and CO mostly increased due to the tar cracking (Eq. 5.5), the

carbon formation (Eq. 5.6), the soot decomposition (Eqs. 5.10 and 5.11), and the steam

75

reforming (Eq. 5.13). CO decreased again as it was converted into CO2 due to radical

termination (Eq. 5.9) and the water-gas shift reaction (Eq. 5.12).

When no steam was fed, the light hydrocarbon gases (CH4, C2H4, and C2H6) were produced at

the concentrations of 0.05%, 0.04%, and 0.06%, respectively. They decreased with the

increase in the steam feed rate, and were hardly generated at the steam feed rate of 0.47 L/min

or higher. Hydrocarbon was generated as benzene was decomposed due to the tar cracking

(Eq. 5.5), and disappeared due to the steam reforming (Eq. 5.13) as steam was fed.

Figure 5.4 Effect of the steam feed rate

▌Effects of Input Benzene Concentration

Figure 5.5 shows the test results when the input benzene concentration was changed from

0.07% to 0.25% to investigate the effects of the aromatic hydrocarbon tar on the pyrolysis

and/or gasification gas.

The benzene decomposition efficiency was 85.2% at the input concentration of 0.07%, and

gradually decreased with the increase in the input concentration. When the input

concentration reached to 0.25%, the decomposition efficiency decreased to 80.7%. Because

the conditions other than the input benzene concentration were fixed as shown in Table 5.3,

almost constant number of electrons and active chemical species, which influence the tar

destruction, were generated. Therefore, with a constant tar destruction capacity, the quantity

of tar that was not decomposed increased due to the increase of input tar.

The destruction energy efficiency significantly increased from 12.5 g/kWh at the input

concentration of 0.07% to 42.1 g/kWh at the input concentration of 0.25%. This was because

the quantity of the decomposed benzene increased, whereas the gas feed rate for the reformer

(Q) and the input electric power (IP) were constant in the destruction energy efficiency

equation (Eq. 5.2).

Carbon-black was completely decomposed due to the soot decomposition (Eqs. 5.10 and

5.11) according to the OH radicals that were generated by steam, when the input benzene

concentration was low. When the input benzene concentration was high at 0.17% or higher,

however, carbon-black started to form. When the input concentration reached its maximum at

0.25%, the carbon-black concentration increased to up to 0.01 g/Nm3. This was because the

input benzene concentration exceeded the decomposition capacity of the OH radicals and

Steam feed rate (L/min)

De

co

mp

ositio

ne

ffic

ien

cy

(%)

De

str

uctio

ne

ne

rgy

eff

icie

ncy

(g/k

Wh

)

Be

nze

ne

co

ncn

etr

atio

n(%

)

0 0.2 0.4 0.6 0.80

20

40

60

80

100

0

5

10

15

20

25

30

0

0.05

0.1

0.15

0.2

0.25

0.3

Output concentration

Decomposition efficiency

Destruction energy efficiency

Input concnetration

(a)


CH

4,C

2H

4,C

2H

6(%

)

0 0.2 0.4 0.6 0.80

0.05

0.1

0.15

0.2

C2H

6

H2

CO

Carbon black

CH4

C2H

4

CO2

H2,C

O,C

O2

(%)

Ca

rbo

nb

lack

(g/N

m3)

0

0.5

1

1.5

2

0

0.02

0.04

0.06

0.08

0.1

(b)

76

more carbon-black was generated due to the carbon formation (Eq. 5.6) remained while a

fixed quantity of steam was supplied.

As for gases, H2 and CO increased with the increase in the input benzene concentration. The

increase rate was higher for H2. This was because benzene was decomposed into H2 due to the

benzene decomposition reactions, which were the tar cracking (Eq. 5.5) and the carbon

formation (Eq. 5.6), and H2 and CO were generated due to the soot decomposition (Eqs. 5.10

and 5.11) and the steam reforming (Eq. 5.13), respectively. In the case of CO2, however, it

slightly increased due to the radical termination (Eq. 5.9), the soot decomposition (Eq. 5.11),

and the water-gas shift reaction (Eq. 5.12). Light hydrocarbon gases (CH4, C2H4, and C2H6)

were hardly generated at the initial stage, as with carbon-black, but started to increase

gradually when the input benzene concentration was 0.17%. This was because the quantity of

hydrocarbon generated due to the tar cracking (Eq. 5.5) was larger than the quantity that was

removed due to the steam reforming (Eq. 5.13), and the generated hydrocarbons were

continuously accumulated.

Figure 5.5 Effect of the input benzene concentration

▌Effects of the Total Gas Feed Rate

Figure 5.6 shows the effects of the total gas feed rate change. The total gas feed rate was set

within the range of 12.6~24.7 L/min, which was a stable state for the discharge of the GAPR.

With the increase in the total gas feed rate, the benzene decomposition efficiency gradually

decreased. This was because the retention time of benzene-containing gas decreased within

the GAPR. Therefore, the reaction time among the electrons, ions, and radicals that were

generated in the plasma discharge and benzene decreased [128].

The destruction energy efficiency gradually increased with the increase of the total gas feed

rate. This is due to the increase in the input benzene concentration with the increase in the

total gas feed rate resulted in the reduction of the tar destruction, but the tar destruction

eventually increased because the gas feed rate (Q) relatively increased, according to Eq. 5.2.

Carbon-black was hardly generated with the increase in the total gas feed rate. This was

because even though the benzene concentration slightly increased with the increase in the

total gas feed rate, the soot was decomposed into light gas (Eqs. 5.10 and 5.11) at a constant

and sufficient steam feed rate of 0.66 L/min.

Input benzene concentration (%)

De

co

mp

ositio

ne

ffic

ien

cy

(%)

De

str

uctio

ne

ne

rgy

eff

icie

ncy

(g/k

Wh

)

Be

nze

ne

co

nce

ntr

atio

n(%

)

0.1 0.15 0.2 0.250

20

40

60

80

100

0

10

20

30

40

50

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7




Input concnetration

(a)

H2,C

O,C

O2

(%)

Ca

rbo

nb

lack

(g/N

m3)

0

0.5

1

1.5

2

2.5

3

0

0.02

0.04

0.06

0.08

0.1

Input benzene concentration (%)

CH

4,C

2H

4,C

2H

6(%

)

0.1 0.15 0.2 0.250

0.05

0.1

0.15

0.2

C2H

6

H2

CO

Carbon black

CH4

C2H

4

CO2

(b)

77

With the increase in the total gas feed rate, H2 and CO gradually increased. H2 increased due

to the benzene decomposition reactions, which were the tar cracking (Eq. 5.5) and the carbon

formation (Eq. 5.6), and H2 and CO increased due to the soot decomposition (Eq. 5.10) and

the steam reforming (Eq. 5.13), respectively, when steam existed. The CO2 did not almost

change because the radical termination (Eq. 5.9), the soot decomposition (Eq. 5.11), and the

water-gas shift reaction (Eq. 5.12) were not predominated. Light hydrocarbon gases (CH4,

C2H4, and C2H6) were not detected. This was because hydrocarbons were converted into H2

and CO due to the steam reforming (Eq. 5.13).

Eventually, the increase in the total gas feed rate reduced the decomposition efficiency due to

a shorter retention time, but because of the characteristic of the test equipment, the input

benzene concentration slightly increased and light gases other than hydrocarbon also

increased.

Figure 5.6 Effect of the total gas feed rate

▌Effects of the Input Electric Energy

Figure 5.7 shows the effects of the specific energy input (SEI).

The benzene decomposition efficiency increased from 82.6% to 92.4% when SEI increased

from 0.17 kWh/m3 to 0.36 kWh/m

3. As the increase in the voltage and current increased the

SEI, the benzene decomposition was accelerated due to the tar cracking (Eq. 5.5). In addition,

when steam existed, the OH radicals increased between the electrodes, and then the benzene

tar was decomposed due to the ring cleavage via their reactions between the OH radicals

[124].

The destruction energy efficiency, however, decreased from 20.9 g/kWh to 10.7 g/kWh

through test range. This was because the input electric power increased, even though the tar

decomposition rate slightly increased with the fixed gas feed rate.

Carbon-black was not detected regardless of the change in the SEI. This was because even

though carbon-black increased due to the carbon formation (Eq. 5.6) with the increase in the

SEI, the increase in the OH radicals completely removed it via the soot decomposition (Eqs.

5.10 and 5.11).

As for the light gases, with the increase in the SEI, H2 and CO increased, but CO2 slightly

decreased. This was because that the plasma tar cracking (Eq. 5.5) of benzene and the carbon


De

co

mp

ositio

ne

ffic

ien

cy

(%)

De

str

uctio

ne

ne

rgy

eff

icie

ncy

(g/k

Wh

)

Be

nze

ne

co

nce

ntr

atio

n(%

)

12 14 16 18 20 22 240

20

40

60

80

100

0

10

20

30

40

50

0

0.05

0.1

0.15

0.2

0.25

0.3




Input concnetration

(a)

H2,C

O,C

O2

(%)

Ca

rbo

nb

lack

(g/N

m3)

0

0.5

1

1.5

2

0

0.02

0.04

0.06

0.08

0.1


CH

4,C

2H

4,C

2H

6(%

)

12 14 16 18 20 22 24

0.05

0.1

0.15

0.2

C2H

6

H2

CO

Carbon black

CH4

C2H

4

CO2

(b)

78

formation (Eq. 5.6) formed light hydrocarbons and carbon, and they were converted into H2,

CO, and CO2 via complex reactions. CO2 decreased due to the CO2 decomposition (Eqs. 5.16

and 5.17), however, more electrons were generated by the plasma discharge with the increase

in the SEI [108].

CO2 + e → CO + O + e- (5.16)

CO2 + e → C+ + O2 + 2e

- (5.17)

The light hydrocarbon gases, CH4, C2H4, and C2H6, increased with the increase in the SEI

because the tar cracking (Eq. 5.5) increased.

This is reasonably supported by the fact that the quantity of the electric transfer between

electrodes increases, i.e., the current increases following an increase in the applied voltage

with the fixed geometry of the electrodes [106].

Figure 5.7 Effect of the input electric energy

▌Effects of the Nozzle Diameter

Figure 5.8 shows the results on the effect of the nozzle diameter. With the parameters in the

optimal condition (Table 5.3), the nozzle diameter was varied as1.5 mm, 3 mm, 4 mm, and 5

mm.

Figure 5.8(a) shows the benzene decomposition efficiency, the energy efficiency, and plasma

discharge pictures. The benzene decomposition efficiency reached its maximum of 84.1%

when the nozzle diameter was 1.5 mm, and decreased to 71.4% when the nozzle diameter

increased to 5 mm. The destruction energy efficiency also decreased from 21.6 g/kWh to 18.6

g/kWh.

As shown in the plasma discharge patterns, when the nozzle diameter was 1.5 mm, the gas

was injected within the 3mm electrode gap and the discharge was uniform over the entire

electrode. When the nozzle diameter was 4 mm or larger, however, the discharge was

insufficient because part of the gas was not injected between the electrodes but was diffused

around them. Especially for the nozzle diameter of 5 mm, the discharge was held at the rear

part of the electrode.

Therefore, for the consecutive plasma discharges in such ranges as the breakdown (A), the

equilibrium heating phase (B), and the non-equilibrium reaction phase (C), as shown in

Specific energy input (kWh/m3)

De

co

mp

ositio

ne

ffic

ien

cy

(%)

De

str

uctio

ne

ne

rgy

eff

icie

ncy

(g/k

Wh

)

Be

nze

ne

co

nce

ntr

atio

n(%

)

0.17 0.2 0.23 0.26 0.29 0.32 0.350

20

40

60

80

100

0

10

20

30

40

50

0

0.05

0.1

0.15

0.2

0.25

0.3




Input concnetration

(a)


CH

4,C

2H

4,C

2H

6(%

)

0.17 0.2 0.23 0.26 0.29 0.32 0.350

0.05

0.1

0.15

0.2

C2H

6

H2

CO

Carbon black

CH4

C2H

4

CO2

(b)

H2,C

O,C

O2

(%)

Ca

rbo

nb

lack

(g/N

m3)

0

0.5

1

1.5

2

0

0.02

0.04

0.06

0.08

0.1

79

Figure 4.1, the gas injection speed must be sufficient to maintain the discharge momentum at

the nozzle.

Figure 5.8(b) shows the concentrations of the light gas. The concentrations of H2, CO, and

CO2 decreased with the increase in the nozzle diameter. This was because that the tar cracking

(Eq. 5.5) and the carbon formation (Eq. 5.6) decreased as a part of the gas deviated from the

plasma range, and the creation of light hydrocarbons and carbon, which are the precursors of

light gas, decreased.

Figure 5.8 Effect of various nozzle diameters

▌Effects of the Electrode Gap

Figure 5.9 shows the results of changing the electrode gap. With the parameters in the optimal

condition, the nozzle gap was varied as 2 mm, 3 mm, and 4 mm.

Figure 5.9(a) shows the benzene decomposition efficiency, the destruction energy efficiency,

and plasma discharge pictures.

The benzene decomposition efficiency was 75.3% when the electrode gap was 2 mm, and it

increased with the increase in the electrode gap. It eventually increased to 87.9% at the

electrode gap of 4 mm. This was because the plasma discharge area increased with the

increase in the electrode gap, and more active species were formed [106, 128]. The

destruction energy efficiency also increased from 18.4 g/kWh to 22.5 g/kWh because the

benzene destruction increased.

As shown in the plasma discharge pictures, when the electrode gap was 2 mm, which was

shorter than the fixed nozzle diameter of 3 mm, not all the gas could be injected in the plasma

discharge column, and part of it was diffused around the electrodes. Therefore, the plasma

discharge was not formed up to the end of the electrode, and the plasma discharge became

relatively unstable. When the electrode gap was 3 mm, the plasma discharge was formed well

in the plasma column, but part of the gas was injected along the plasma column boundary, and

the plasma discharge was unstably formed beyond the electrode blade in a fluttering manner.

When the electrode gap was wide at 4 mm, the gas was properly injected within the plasma

column, and the discharge stabilized. The volume of the plasma column was also larger than

in the aforementioned cases.

Ga

sco

nce

ntr

atio

n(%

)

0

0.2

0.4

0.6

0.8

1

CH4

H2

CO CO2

C2H

4C

2H

6

5.0 mm

1.5mm

3.0 mm

4.0 mm

Nozzle diameter (b)

80

Figure 5.9(b) shows the concentrations of the light gases. H2, CO, and CO2 gases were formed.

Each gas increased with the electrode gap due to the increasing of the discharge column and

the improved stabilization. All the carbon-black and hydrocarbon might be converted to the

light gas.

It is known that in the GAPR design, the nozzle diameter and the electrode gap are important

factors that influence the plasma discharge pattern, and that the stable plasma can be obtained

when the ratio of the electrode gap to the nozzle diameter is 1 or higher [128].

Figure 5.9 Effect of various electrode gaps

▌Effects of the Electrode Length

Figure 5.10 shows the test results of changing the electrode length. With the parameters in the

optimal condition, the electrode length was varied as 70 mm, 95 mm, and 125 mm.

Figure 5.10 Effect of various electrode lengths

Ga

sco

nce

ntr

atio

n(%

)

0

0.2

0.4

0.6

0.8

1

H2

CO CH4

C2H

4C

2H

6

4 mm

Electrode gap

3 mm

2 mm

CO2

(b)

Ga

sco

nce

ntr

atio

n(%

)

0

0.2

0.4

0.6

0.8

1

1.2

COH2

CO2

CH4

C2H

4C

2H

6

Electrode length

70 mm

95 mm

125 mm

(b)

81

The benzene decomposition efficiency increased from 78.7% to 87.9% as the electrode length

increased from 70 mm to 125 mm. The destruction energy efficiency also increased from 18.4

g/kWh to 22.5 g/kWh. As shown in the plasma discharge pictures, a longer electrode led to a

larger discharge range, and the benzene decomposition efficiency and the destruction energy

efficiency increased due to the increase in the gas retention time.

Carbon-black was not formed. Light gases (H2, CO, and CO2) were formed, and they

increased with the increase in the electrode length.

▌Effects of the Electrode Shape

Figure 5.11 shows the results of changing the electrode shapes with the parameters in the

optimal condition.

With the triangle-shaped electrode, the plasma discharge was wider, but the plasma discharge

was not formed from the breakdown starting point to the peak of the electrode. The plasma

discharge disappeared in the middle of the electrode. With the Arc 1 electrode, the plasma

discharge was formed from the breakdown starting point to the peak of the electrode, and the

plasma discharge was formed over the entire electrode [129]. However, the Arc 2 and Arc 3

electrodes had different points at which the discharge diverged. The increase in the distance

between the breakdown point and the plasma discharge diversion point reduced the volume of

the plasma discharge range. Therefore, Arc 1 had the best plasma discharge volume and

stability, and the benzene decomposition efficiency and the destruction energy efficiency

were highest.

Carbon-black was not formed. Light gases (H2, CO, and CO2) were formed, in the descending

order of Arc 1, Arc 2, Arc 3 and the triangle.

Figure 5.11 Effect of various electrode shapes

2) Destruction for light PAH tar

As a respresentative light PAH (polycyclic aromatic hydrocarbon) tar, anthracene was

selected. Test was performed at the optimal conditions for maximizing the anthracene

Ga

sco

nce

ntr

atio

n(%

)

0

0.2

0.4

0.6

0.8

1

CH4

H2

CO CO2

C2H

4C

2H

6

Arc3

Triangle

Arc1

Arc2

Electrode shape (b)

82

decomposition and the destruction energy efficiency of the GAPR to verify the anthracene tar

destruction. Table 5.4 shows the optimal operating conditions and results.

At the optimal condition, the anthracene decomposition efficiency was 96.1%, and the

destruction energy efficiency was 1.14 g/kWh. The higher heating value of the gas produced

by the anthracene decomposition (steam reforming) was 11,324 kJ/Nm3. The carbon balance

was 98%. It seems that the value of the carbon balance did not reach 100% because some of

the deposited and created carbon-black was converted into HCN in the reactor [123]. The CH

radicals which are produced by the ring cleavage in tars and in some intermediates, forms the

HCN and CN radical (Refer to Eqs. 5.14 and 5.15).


Optimal conditions

Conditions Steam feed rate

(L/min)

Input tar

concentration (g/m3)

Total gas feed rate

(L/min)

Specific energy input

(kWh/m3)

Value 0.63 0.21 12.05 0.175

Experiment results

Result

Gas composition after the reformer

(%, N2 excluded) Carbon

black

(g/Nm3)

Carbon

balance

(%)

Higher

heating

value

(kJ/Nm3)

Decomposition

efficiency

(%)

Destruction

energy

efficiency


79.2 9.5 11.3 0 0 0 0 98 11,324 96.1 1.14

In addition, this study was performed by changing the parameters which affect the

decomposition and the destruction energy efficiency. The tests were performed according to

the parameter ranges in Table 5.2. The other parameters were fixed at the optimum values as

shown in Table 5.4.

▌Effects of the Steam Feed Rate

Figure 5.12 shows the results of changing the the steam feed rate with the total gas feed rate

of 12.05 L/min and the SEI of 0.17 kWh/m3. When the steam feed rate exceeded 1.57 L/min,

the temperature of the steam generator started to decrease. Therefore, the testing range of the

steam feed rate was determined to be 0~1.57 L/min.

The decomposition efficiency was 61% without steam feed (i.e., the steam feed rate of 0

L/min). The decomposition efficiency increased with the increase in the steam feed rate, and

it reached to 96.1% at the steam feed rate of 0.63 L/min. Then the decomposition decreased

with the increase in the steam feed rate. When steam was not fed, the tar cracking reaction

(Eq. 5.5) decomposed tar to create hydrocarbons and hydrogen. In addition, the tar generated

carbon-black and hydrogen according to Eq. 5.6 [26].

Thereafter, the steam that was fed into the GAPR produced OH radicals according to the

radical production reaction (Eq. 5.7). As shown in Eq. 5.8, the created OH radicals reacted

with tar and converted it into other products [117]. Accordingly, the tar decomposition

efficiency increased with the steam injection. Temperature in the GAPR was 250℃ for the

cold air plasma. But in the case of hot steam feeding and line heating in this study, the

83

temperature in the GAPR was higher than the cold air plasma (e.g. 380℃ at the optimal

condition). So, the tar destruction might be affected slightly by the production of the OH

radical, the reaction rate, etc.

However, the steam amount also has an adverse effect on tar removal due to its

electronegative characteristics [120]. Too many water molecules caused by the increase of the

steam feed amount limits the electron density in the GAPR and quench the activated chemical

species. That is why the decomposition efficiency decreased after reaching the maximum

value.

The destruction energy efficiency decreased gradually with the increase of the steam feed rate.

Increasing the steam feed amount results in brings about lower input tar concentrations so that

the tar removal expressed by Eq. 5.2 was decreased. That is why the destruction energy

efficiency decreased.

The carbon-black concentration was 0.51 g/Nm3 without the steam injection. By injection

steam, it decreased significantly, showing almost zero value at the steam feed rate of 0.37

L/min or more. Carbon-black formed by the reaction of Eq. 5.6, was decomposed by the soot

decomposition reactions (Eqs. 5.10 and 5.11), where carbon-black was oxidized to CO, CO2,

and H2 due to the OH radicals [124].

The major reformed gases included H2, CO, and CO2. Small quantities of light hydrocarbon

gases (CH4, C2H4, and C2H6) were also observed. H2 continued to increase according to Eqs.

5.5, 5.6, 5.10, and 5.11, and reached to the concentration of 0.77%. The CO concentration

increased to 0.09% until the steam feed rate reached to 0.63 L/min according to Eq. 5.10, but

it decreased according to the water-gas shifting reaction (Eq. 5.12) when the steam feed rate

exceeded 0.63 L/min. CO2 was created according to Eq. 5.11 and increased slightly according

to Eq. 5.12 [126].

The light hydrocarbon gases decreased with the increase in the steam feed rate, and they were

not produced at the steam feed rate of 0.5 L/min or more. The tar was decomposed to make

hydrocarbon substances according to Eq. 5.5. With the increase in the steam feed rate, the

hydrocarbons were converted into H2 and CO according to the steam reforming reaction (Eq.

5.13) [125].



An

thra

ce

ne

&C

arb

on

bla

ck

(g/N

m3)

De

co

mp

ositio

ne

ffic

ien

cy

(%)

De

str

uctio

ne

ne

rgy

eff

icie

ncy

(g/k

Wh

)

0 0.3 0.6 0.9 1.2 1.50

0.2

0.4

0.6

0.8

1

0

20

40

60

80

100

0

0.5

1

1.5

2

2.5

3

Carbon black conc.


Input concentration



CH

4,C

2H

4,C

2H

6(%

)

0

0.05

0.1


H2,C

O,C

O2

(%)

0 0.3 0.6 0.9 1.2 1.50

0.3

0.6

0.9

1.2

CO

Steam feedrate (L/min)

C2H

6(ppm)

CO2

0.5

C2H

4(ppm)

H2

CH4

C2H

4

C2H

6

CH4

(ppm)

0

0.06 0.18 0.31 0.40

212 211 19 0 0

109 92 53 0 0 0

834 328 147 43 38 17

0.6

0

0

0

84

▌Effects of the Input Anthracene Concentration

Figure 5.13 shows the effect of the anthracene input concentration. The test was conducted at

the input anthracene concentration of 0.1~0.7 g/Nm3.

The decomposition efficiency decreased with the increase in the input anthracene

concentration. Particularly, at the input anthracene concentration of 0.36 g/Nm3 or more, the

decomposition efficiency had a lower value than about 80% which cannot be accepted as the

reformer. The reason is that the amounts of electrons and active species from the plasma

discharge were constant due to the excess of designed capacity in the GAPR.

The change in the input anthracene concentration significantly influenced the destruction

energy efficiency. The destruction energy efficiency proportionally increased with the

increasing of the input anthracene concentration. This is because the anthracene removal

increased whereas the decomposition efficiency decreased.

Carbon-black increased slightly at the tar concentration of 0.27 g/Nm3 or more. The reason is

that the fixed steam feed amount could not produce enough OH radical to react with

anthracene (Eq. 5.8) or carbon (Eqs. 5.10 and 5.11).

With the increase in the input anthracene concentration, H2 significantly increased, while CO,

CO2, CH4, C2H4, and C2H6 slightly increased. The amount of the increased anthracene gives

the conversion to higher light hydrocarbon. The light hydrocarbons react with each gas

according to Eqs. 5.5, 5.6 and 5.10~5.13, respectively. That is because the light gases

increased with the increasing of the anthracene concentration.

Figure. 5.13 Effects of the input anthracene concentration


Figure 5.14 shows the total gas feed rate change. The total gas feed rate was controlled and

kept within the range of 7.2~30.1 L/min.

The discharge was unstable at the total gas feed rate of 7 L/min or below, because the gas

velocity at the exit of the nozzle was low. And at the total gas feed rate of 30 L/min or more,

the plasma discharge blew off due to high gas velocity. Therefore, the test range of the total

gas feed rate was determined to be 7.2~30.1 L/min.

The decomposition efficiency slightly increased and it then had the maximum value of 88.5%

at 12.05 L/min due to the best plasma discharge. After reaching that value, the efficiency

Input anthracene concentration (g/Nm3)

An

thra

ce

ne

&C

arb

on

bla

ck

(g/N

m3)

De

co

mp

ositio

ne

ffic

ien

cy

(%)

De

str

uctio

ne

ne

rgy

eff

icie

ncy

(g/k

Wh

)

0.1 0.2 0.3 0.4 0.5 0.6 0.70

0.2

0.4

0.6

0.8

1

0

20

40

60

80

100

0

0.5

1

1.5

2

2.5

3

Carbon black conc.



Input concentration


CH

4,C

2H

4,C

2H

6(%

)

0

0.05

0.1

Input anthracene concentration (g/Nm3)

H2,C

O,C

O2

(%)

0.1 0.2 0.3 0.4 0.5 0.6 0.70

0.3

0.6

0.9

1.2

CO

H2

CH4

CO2

C2H

4

C2H

6

85

decreased gradually. This tendency is possibly due to the shortenings of both the contact area

and the interaction time between anthracene tar and other reactants in the plasma discharge

zone, which leads to the reductions of energetic electrons impact dissociation and also the

reactions between tar and reactive ions and OH radicals [128].

The destruction energy efficiency increased significantly up to the total gas feed rate of 24

L/min (2.63 g/kWh), having almost the constant value after that. This was because the tar

removal decreased due to the decrease in the retention time.

Carbon-black was not collected regardless of the change of the total gas feed rate because the

amount of the steam feed was fixed at 0.37 L/min. The fed steam creates OH radicals, which

can oxidize the generated carbon according to the reactions shown by Eqs. 5.10 and 5.11.

With the increase of the total gas amount, H2 increased gradually. This is because the

secondary gases reacted significantly with high gas interactions (Eqs. 5.10~5.13). CO

decreased with the increase in the total gas feed rate, while CO2 increased. This was because

the carbon that was produced from the initial cracking reaction was converted into CO

according to Eq. 5.10, and the produced CO was converted into CO2 according to Eq. 5.12.

C2H4 and C2H6 were not generated, but CH4 increased with the increase in the total gas feed

rate. This is due to the decomposition of C2H4 and C2H6 into CH4. Some of the typical

reactions are shown in Eqs. 5.18 and 5.19 [89]:

C2H6 → C2H4 + H2 (5.18)

C2H4 → CH4 + C (5.19)

Figure 5.14 Effect of the total gas feed rate

▌Effects of the Input Electric Energy

Figure 5.15 shows the effect of the specific energy input (SEI) (refer to Eq. 5.3). The

experiments were conducted within the SEI range of 0.175~0.234 kWh/m3.

With the increase in the SEI, the decomposition efficiency increased gradually. The

decomposition efficiency was 88% at the SEI of 0.175 kWh/m3, and it increased to 94.1% at

the SEI of 0.234 kWh/m3. As the increase of the SEI, the electrons and OH radicals between

the electrodes increased. The created electrons and OH radicals became more active due to the

increased input electric power [106]. And then the radicals converted tar to product gases (H2,

CO, and CO2), showing increasing tar destruction.


An

thra

ce

ne

&C

arb

on

bla

ck

(g/N

m3)

De

co

mp

ositio

ne

ffic

ien

cy

(%)

De

str

uctio

ne

ne

rgy

eff

icie

ncy

(g/k

Wh

)

5 10 15 20 25 300

0.2

0.4

0.6

0.8

1

0

20

40

60

80

100

0

0.5

1

1.5

2

2.5

3

Carbon black conc.

Input concentration




CH

4,C

2H

4,C

2H

6(%

)

0

0.05

0.1


H2,C

O,C

O2

(%)

5 10 15 20 25 300

0.3

0.6

0.9

1.2

CO

H2

CH4

CO2

C2H

4

C2H

6

Total gas feedrate (L/min)

CH4

(ppm)

30.1

59

8.4 12 14.5 18.1 24.17.2

13 16 36 39 44 48

86

Figure 5.15 Effect of the input electric energy

Carbon-black was not collected within the test range because carbon-black converted to

producer gases by OH radical as shown in Eqs. 5.10 and 5.11.

H2 increased significantly by increasing in the SEI up to 0.234 kWh/m3 according to Eqs. 5.5,

5.6 and 5.10~5.13. CO and CO2 increased slightly according to Eqs. 5.10~5.13, having lower

values than H2. Hydrocarbons (CH4, C2H4, C2H6) increased slightly.

Table 5.5 represents the comparison to other researches for a surrogate tar. The results of the

benzene tar (Table 5.3) and anthracene tar (Table 5.4) were taken at the optimal conditions.

The decomposition efficiency (Eq. 5.1) and destruction energy efficiency (Eq. 5.2) should be

affected by the plasma discharge type and the model tar type.

Table 5.5 Comparison between this study and other researches for decomposition of the

surrogate light tar

Researches This study Yu et al. [123] Tippayawong et

al. [26] Du et al. [120]

Plasma discharge type 3 phase AC gliding arc DC gliding arc AC gliding arc AC gliding arc

Light tar model Benzene Anthracene Naphthalene Naphthalene Toluene


(%) 82.6 96.1 92.3 95 98.5

Destruction energy

efficiency (g/kWh) 20.9 1.14 3.6 0.123 29.46

Gas flow rate (L/min) 16.7 12.05 6.8 9.16 13.3

Input power (kW) 0.152 0.128 0.12 0.55 0.209

SEI (kWh/m3) 0.17 0.175 0.4686 1 0.26

Tar removal (g/m3) 3.17 0.202 1.219 0.1235 7.73

Input concentration 0.12%

(3.83 g/m3)

0.21 g/m3 1.32 g/m

3 0.13 g/m

3 7.85 g/m

3

Output concentration 0.02%

(0.66 g/m3)

0.008 g/m3 0.101 g/m

3 0.0065 g/m

3 0.118 g/m

3


An

thra

ce

ne

&C

arb

on

bla

ck

(g/N

m3)

De

co

mp

ositio

ne

ffic

ien

cy

(%)

De

str

uctio

ne

ne

rgy

eff

icie

ncy

(g/k

Wh

)

0.175 0.185 0.195 0.205 0.215 0.225 0.2350

0.2

0.4

0.6

0.8

1

0

20

40

60

80

100

0

0.5

1

1.5

2

2.5

3

Carbon black conc.

Input concentration




CH

4,C

2H

4,C

2H

6(%

)

0

0.05

0.1


H2,C

O,C

O2

(%)

0.9 0.95 1 1.05 1.1 1.15 1.20

0.3

0.6

0.9

1.2

CO

H2

CH4

CO2

C2H

4

C2H

6

87

Although lower input tar concentration like in this study should be hard to be removed, the

decomposition efficiency was 82.6% for benzene and 96.1% for anthracene, respectively. In

case of the benzene tar, the decomposition efficiency was little lower value because the SEI

(Eq. 5.3) was set to low value, compared to other researches having higher efficiency.

The destruction energy efficiency for the benzene and anthracene decomposition in this study

had the values of 20.9 g/kWh and 1.14 g/kWh, respectively.

The destruction energy efficiency should be mainly affected by the tar removal which has

high value for larger amount of the input concentration and by the gas flow rate. That is why

the destruction energy efficiency of the benzene showed higher than the anthracene and the

values reported by Yu et al. [123] and Tippayawong et al. [26].

In conclusion, this work used the benzene and anthracene tars with lower input concentrations

which is harder situation than other researches. Nevertheless, the results showed generally

better due to using the 3 phase AC gliding arc plasma.

5.3.2 Verification of tar removal in the continuous pyrolyzer

1) Test setup for tar removal in the biomass pyrolysis

Figure 5.16 exhibits the test equipment diagram that was designed to verify the real tar

removal performance of the GAPR. Tar is a pyrolysis product from a wood chip biomass fuel.

The temperature of the pyrolyzer was controlled by the electro-furnace, and the wood chips

were supplied by a screw pyrolyzer. The producer gas from the wood chips pyrolysis was

carried to the GAPR by a nitrogen carrier gas, and a steam generator whose temperature was

set at 300℃ to produce steam for the reforming process. The generated steam was fed along

with the carrier gas, which is required for a stable plasma discharge, to the GAPR. The steam

feed rate was 0.3 L/min.

The wet-type tar sampling and analysis of Biomass Technology Groups (BTGs) were

conducted for this test [82]. The gravimetric tar mass was determined to measure the

pyrolysis gas products and the tar yields after the reforming. Benzene, naphthalene,

anthracene, pyrene, benzonitrile, and benzoacetonitril were analyzed to determine the

concentrations of representative light tar components.

Immediately after completing the sampling, the contents of the impinger bottles were filtered

through a filter paper (Model F-5B, Advantec Co., Japan). The filtered isopropanol solution

was divided into two parts. The first was used to determine the gravimetric tar mass by means

of the solvent distillation and evaporation with an evaporator (Model N-1000-SW, Eyela,

Japan), in which the temperature and the vapor pressure were 55~57℃ and 230 hPa,

respectively.

The second was used to determine the concentrations of light tar compounds using the GC-

FID (Model 14B, Shimadzu, Japan). Quantitative tar analysis was performed by the GC

system, using a RTX-5 (RESTEK) capillary column (30 m-0.53 mm id, 0.5 μm film

thickness).

The syngas produced by reforming in the GAPR was analyzed with the GC-TCD (Model CP-

4900, Varian, Netherlands). Molecular Sieve 5A columns were used for H2, CO, O2, and N2

analysis, and poraPlot-Q columns for CO2, C2H4, and C2H6 analysis.

88

Figure 5.16 Experimental setup for biomass tar removal in a plasma reformer

2) Experimental results of the decomposition of biomass tar by the plasma

reformer

An experiment was conducted to verify the tar removal performance of the GAPR, which was

connected to the back of the continuous-type screw pyrolyzer, using wood chips as a biomass

fuel. For the plasma conditions, the steam feed rate and the input electric power (SEI) were

stably maintained at 0.3 L/min and 0.91 kWh/m3, respectively, in the experiment.

Figure 5.17 shows the gravimetric tar mass and the concentrations of the selected light tar

before and after the GAPR. The gravimetric tar mass was significantly reduced to 3.74 g/Nm3

at the outlet of the GAPR (from 18.02 g/Nm3 at the inlet of the reformer). The removal

efficiency was 79.2%, accordingly.

The concentrations of the light tar compounds were also significantly reduced to 0.46 g/Nm3

from 3.47 g/Nm3 for benzene, 0.11 g/Nm

3 from 0.37 g/Nm

3 for naphthalene, 0.03 g/Nm

3 from

0.09 g/Nm3 for anthracene, 0.02 g/Nm

3 from 0.07 g/Nm


3 from 0.85

g/Nm3 for benzonitrile, and 0.0 g/Nm

3 from 0.04 g/Nm

3 for benzoacetonitril, after the

reforming process.

The decomposition of heavy tar occurred due to the tar cracking (Eq. 5.5) and the carbon

formation (Eq. 5.6). The steam feed to the GAPR produced water excitation species, as

presented in Eq. 5.20, using a plasma discharge. Thus, the light tar and carbon produced

during the decomposition of heavy tar were converted to light gases [130].

H2O → H, e−, OH, H2, H2O2, H3O

+, OH (5.20)

89

Figure 5.17 Light tar contribution before and after the plasma reformer

Figure 5.18 shows the change in the light gas concentrations after the reforming process. The

light gas composition shifted as a result of the combined reactions of the gas products from

the tar decomposition and the pyrolysis gas.

Figure 5.18 Light gas concentrations before and after the plasma reformer

The concentrations of the pyrolysis gas components were 20.3% H2, 43.9% CO, 13.6% CO2,

16.4% CH4, 5.1% C2H4, and 0.4% C2H6 at the inlet of the GAPR, but these converted to

42.1% H2, 34.0% CO, 8.7% CO2, 3.0% CH4, 1.2% C2H4, and 0.1% C2H6 at the outlet of the

GAPR. That is, H2 and CO2 increased due to the tar cracking (Eq. 5.5), the carbon formation

(Eq. 5.6) and the soot decomposition (Eqs. 5.10 and 5.11), while CO and light hydrocarbons

(CH4, C2H4, and C2H6) decreased by the water-gas shift reaction (Eq. 5.12) and the steam

reforming (Eq. 5.13) after the plasma process.

Gra

vim

etr

icta

r(g

/Nm

3)

Se

lecte

dlig

ht

tar

(g/N

m3)

0

5

10

15

20

0

0.5

1

1.5

2

2.5

3

3.5

4

Screw pyrolyzer

GAPR

Gravimetrictar

Anthra-cene



Benzene Naph-thalene

Ga

sco

nce

ntr

atio

n(%

)

0

10

20

30

40

50

60

Screw pyrolyzer

GAPR

H2

CO CO2

CH4

C2H

6C

2H

4

90

5.3.3 Plasma reformer with an external oscillation

1) Test setup and procedure for tar destruction

For enhancing the idea of the plasma reformer in the future, an externally oscillated plasma

reformer (EOPR) was designed and verified for its performance to destruct tar. (The merits of

EOPR should be explained) Benzene was used as the representative tar substance.

The test rig shown in Figure 5.19 consisted of an EOPR, an oscillation control device, a steam

feeding line, a tar feeding line, a power supply equipment, a measurement-analysis line, and a

control-monitoring system.

The oscillation control device included a loud speaker (BT40, Speaker Mall, South Korea), an

amplifier (PA-4000A, INTER-M, South Korea), and a function generator (Agilent 33250A,

Agilent Technology, USA). At the rear of the EOPR, a sound pressure level meter (DSL-330,

TECPEL, Taiwan) was installed to measure the sound pressure in the EOPR.

Other parts in the experimental setup are shown in Figure 5.1.

Figure 5.19 Schematic diagram of the test setup for an EOPR

The test was performed with the parameters that influence the tar decomposition and the

destruction energy efficiency, such as the oscillation frequency, the oscillation amplitude, the

steam feed rate, and the total gas feed rate. Table 5.6 shows the test range for these parameters.

Table 5.6 Test conditions and range for each parameter

Experimental

conditions

Oscillation frequency

(Hz)

Oscillation

amplitude (Vpp)

Steam feed rate

(L/min)

Total gas feed rate

(L/min)

Range 0~1000 0.5~3 0~0.85 12.4~24.5

91

Eq. 5.21 shows the definition of the sound pressure.

20

(dB) L

oS

p

10p (Pa) P (5.21)

where SP

is the sound pressure ( Pa ),

op is the reference values (threshold of hearing)

( Pa 102Pa20 -5 ) and Lp is the sound pressure level (dB).

▌Acoustic wave interaction with the plasma discharge

An increase in the acoustic wave frequency implies that a perturbation producing rarefaction

and compression in the plasma has a small wave length, thus causing the plasma particles to

collide with an increasing rate. The acoustic wave when passing through a sufficiently dense

plasma may travel with a supersonic speed as a shock wave.

The variation of the electron density (α) with the acoustic wave frequency (ω) may be

explained as follows. When the acoustic wave frequency (ω) is less than the electron-atom

elastic collision frequency (ν), the rate of variation of pressure in the plasma discharge is low

enough such that the change in the electron density can easily follow the pressure perturbation.

Figure 5.20 The variation of the electron density and the collision frequency as a function of

the acoustic wave frequency

When the acoustic wave frequency (ω) is higher than the electron-atom elastic collision

frequency (ν), the electron velocity is no longer able to follow the variations of the pressure

distributions, and hence the electron density decreases. The collision frequency (β) depends

on the pressure variations; an increase in ω causes the plasma particle to collide rapidly, thus

increasing β [131].

2) Experimental results of the benzene tar decomposition by the EOPR

An EOPR was developed to destruct tar from the pyrolysis and/or gasification of organic

waste resources, biomass, etc. Benzene was selected as tar representative tar compound, and

the test was performed by changing the various parameters. Table 5.7 shows the test results

under the optimal conditions, showing the maximum tar decomposition and the destruction

energy efficiency.

92


Experimental conditions for each parameter

Condition

s

Oscillation

frequency

(Hz)

Oscillation

amplitude

(Vpp)

Steam feed

rate

(L/min)

Total gas

feed rate

(L/min)

Specific

energy input

(kWhm-3

)

Input

benzene

conc. (%)

Value 267 3 0.66 16.4 0.17 0.12

Experiment results

Result

Reforming gas (%, N2 excluded) Carbon

black

(g/Nm3)

Carbon

balance

(%)

Higher

heating

value

(kJ/Nm3)

Decom-

position

efficiency

(%)

Destruction

energy

efficiency


With

oscillation 39.2 37.1 23.7 0 0 0 0 88.8 9,718 90.7 23.0

Without

oscillation 38.9 33.4 27.6 0 0 0 0 91.4 9,209 82.6 20.9

With the external oscillation, the benzene decomposition efficiency was 90.7%, and the

destruction energy efficiency was 22.95 g/kWh. The light gases that were produced from the

benzene decomposition included H2, CO, and CO2. The higher heating value was 9,718

kJ/Nm3. The carbon balance was 82.5%. It seems that the value of the carbon balance did not

reach 100% due to the heavy hydrocarbons, the nitric tar products (HCN and CN), carbon

black, etc. from the benzene conversion products, were not considered [123]. Without

oscillation, the decomposition efficiency was 82.6%, the destruction energy efficiency was

20.9 g/kWh, and the higher heating value was 9,209 kJ/Nm3, which were smaller than those

with the external oscillation.

Figure 5.21 shows the plasma discharge with and without oscillation under the optimal

condition. Sound wave gives spatio-temporal variations of the gas pressure due to the

expansion and compressions of gas. So, the gas discharge should be influenced by the sound

wave irradiation. Irradiating the sound wave and increasing the sound pressure, the luminous

part spreads wider due to the vibration of the gaseous medium. The expansion of the streamer

was probably caused by a cyclic change in the discharge field due to the violent vibration of

medium particles in the neighborhood of the electrodes.

(a) Without oscillation (b) With oscillation

Figure 5.21 Photo of plasma discharge with and without oscillation

93

On the left, the plasma discharge was not forced acoustically; On the right, its instability was

forced by means of periodic sound waves introduced through a loudspeaker near the plasma

discharge at its natural frequency. The forced acoustic waves reduce the length of the laminar

boundary layer on the periphery of the plasma discharge and cause more regular formation of

vortex rings than under the unforced [132].

Figure 5.22 shows the sound pressure (Ps) as a function of the oscillation frequency at the

fixed oscillation amplitude of 3 Vpp under the optimal condition. The sound pressure was

calculated with Eq. 5.21 by using the sound pressure level (Lp) measured from the EOPR

outlet with external oscillation.

External oscillation to the plasma discharge causes expansion of the discharge space, and the

degree of the expansion depends on the magnitude of the irradiated sound pressure. Standing

sound waves are formed due to the interference of the incident and reflected sound waves

inside the EOPR. Under the standing sound wave field, the sound pressure and the particle

velocity, which is defined as the vibration velocity of the gaseous medium due to the sound

wave, is distributed in the acoustic tube [133].

Figure 5.22 Sound pressure according to the oscillation frequency

Resonance states were observed with the external oscillation frequency at 267, 560, and 810

Hz. The sound pressure (Ps) at the end of the EOPR tube is proportional to the particle

velocity at the loop of the distribution.

▌Effects of the Oscillation Frequency

The factors governing the tar destruction rate in the plasma discharge can be described by Eq.

(5.22) [123, 134].

ned NNkVr (5.22)

where k is the reaction rate constant, dV is the volume of the space where discharge takes

place, eN is the density of discharged electrons, nN is the density of the gas in the space,

and is the formation frequency of active species.

Oscillation frequency (Hz)

Ps

(Pa

)

100 200 300 400 500 600 700 800 900 10000

10

20

30

40

50

16 Pa(267 Hz)

43 Pa(560 Hz)

23 Pa(810 Hz)

94

Figure 5.23 shows the test results at the oscillation frequencies ranging from 0 to 1,000 Hz,

with the parameters fixed at the optimal conditions (Table 5.6).

Figure 5.23(a) shows the benzene decomposition and the energy efficiencies as well as the

benzene concentrations at the inlet and outlet of the EOPR.

The decomposition efficiency increased, and after having a peak value it decreased. The

major factor for the tar benzene destruction rate is the acoustic wave frequency (Ø ) in the

EOPR as shown in Eq. 5.22. The resonance states were observed with the external oscillation

frequency at 267, 560, and 810 Hz as already explained in Figure 5.22. Especially, the

oscillation frequency at 267 Hz might impose fewer constraints which means THE greatest

effect although the frequencies at 560 and 810 Hz have higher sound pressures. Therefore, the

increase in the external oscillation frequency increased the tar destruction up to 267 Hz which

has the maximum benzene decomposition efficiency of 91%. This is because the sound

pressure causes expansion of the discharge space in the plasma discharge.

But when the acoustic wave frequency (ω) is higher than the electron-atom elastic collision

frequency (ν), the electron velocity is no longer able to follow the variations of the pressure

distributions, and hence the electron density (α) decreases as already explained in Figure 5.20.

Therefore, the decrease in the destruction rate due to the sound-wave irradiation can be

attributed to the negative effects of the decrease in the electron density, although having the

positive effects of the increase in the collision frequency. So, after having a peak value, the

decomposition efficiency was decreased.

The destruction energy efficiency had a similar tendency to the decomposition efficiency.

This is because the main factor affecting the destruction energy efficiency (calculated by Eq.

5.2) is the tar removal which is the difference between the input and output concentration.

Figure 5.23(b) shows the concentrations of light gas and carbon-black. The light gases

produced were H2, CO, and CO2, having low concentrations due to a low tar input.

Particularly, H2 and CO had highest values at 267 Hz. But CH4, C2H4, C2H6 and carbon-black

were not almost detected.

Figure 5.23 Effects of the oscillation frequency


De

co

mp

ositio

ne

ffic

ien

cy

(%)

De

str

uctio

ne

ne

rgy

eff

icie

ncy

(g/k

Wh

)

Be

nze

ne

co

nce

ntr

atio

n(%

)

0 200 400 600 800 10000

20

40

60

80

100

0

5

10

15

20

25

30

0

0.05

0.1

0.15

0.2

0.25

0.3




Input concnetration

(a)


CH

4,C

2H

4,C

2H

6(%

)

0 200 400 600 800 10000

0.05

0.1

0.15

0.2

C2H

6

H2

CO

Carbon black

CH4

C2H

4

CO2

(b)

H2,C

O,C

O2

(%)

Ca

rbo

nb

lack

(g/N

m3)

0

0.5

1

1.5

2

0

0.02

0.04

0.06

0.08

0.1

95

Figure 5.24 shows the plasma discharge by changing the oscillation frequency. At the

frequency of 267 Hz, when the benzene decomposition and the energy efficiencies were

highest, the plasma discharge was the most active, although it was difficult to accurately

identify.

0 Hz 267 Hz 560 Hz 810 Hz

Figure 5.24 Photos of the plasma discharge for various oscillation frequencies

▌Effects of the Sound Pressure

Figure 5.25 shows the test results with the oscillation amplitude varied within the 0.5~3 Vpp

range.

Figure 5.25(a) shows the decomposition and energy efficiencies as well as the benzene

concentrations at the inlet and outlet of the EOPR.

The decomposition efficiency gradually increased with increasing the oscillation amplitude.

The change in the oscillation amplitude influences the gas pressure. The concentration of

heavy particles (neutral atoms, positive or negative ions) increases with increasing the

oscillation amplitude [135]. So, the heavy particles react with each other to destruct benzene.

A number of papers have reported the observation of an increase in the pressure amplitude A

of sound waves in gas plasma discharges in comparison with un-ionized air at identical values

of the static gas pressure P0, amplitude ξ of the displacement of the gas particles in the sound

wave, and the sound frequency ν [136]:

A = 2π νP0 ξ γ / W (5.23)

where W is the sound velocity and γ is the ratio of the specific heats.

The destruction energy efficiency showed a trend similar to that of the decomposition

efficiency. This was because the destruction energy efficiency was influenced by the

decomposition efficiency when the gas feed rate in the plasma reformer (Q) and the input

electric energy (IP) were constant.

Figure 5.25(b) shows the concentrations of light gases and carbon black, which were

produced from the tar benzene decomposition.

The light gases produced were H2, CO, and CO2, having low concentrations due to a low tar

input. The concentrations of H2 and CO2 were almost constant (0.82 and 0.48% on average,

respectively) regardless of the change in the sound pressure. The CO concentration slightly

increased from 0.65 to 0.82% with the increase in the sound pressure. But hydrocarbons (CH4,

C2H4, and C2H6) and carbon-black were not detected.

96

Figure 5.25 Effects of the oscillation amplitude

▌Effects of the Steam Feed Rate

Figure 5.26 shows the test results with the steam feed rate varied within the 0~0.85 L/min

range.


concentrations at the inlet and outlet of the EOPR, with and without oscillation.

With the increase in the steam feed rate, the decomposition efficiency gradually increased and

reached 90.7% at the steam feed rate of 0.66 L/min. Then it started to decrease.

When the steam feed rate was 0 L/min, that is, when no steam was supplied, the

decomposition efficiency was 77.4%. This was because the tar was decomposed according to

the tar cracking (Eq. 5.5) and the external-oscillation effect (Eq. 5.22), without the effect of

steam.

As steam was supplied, OH radicals, electrons, and active chemical species were created due

to the water excitation (Eq. 5.7). Then the benzene tar was decomposed due to the radical

utilization (Eq. 5.8), and then the decomposition efficiency increased.

However, water also has an adverse effect on tar removal due to its electronegative

characteristics. Too many water molecules limits the electron density in the plasma discharge

in the EOPR and quench the activated chemical species [120]. Therefore, after having

maximum value, the decomposition efficiency decreased due to too much feeding of steam. In

addition, with the increase in the steam feed rate, the total gas feed rate in the EOPR increased.

Accordingly, a sufficient retention time was not ensured, and the decomposition efficiency

decreased.

Without oscillation, the decomposition efficiency according to the change in the steam feed

rate showed almost the same pattern as that with oscillation. The decomposition efficiency

was higher, however, with external oscillation. This was because the OH radicals, electrons,

and active chemical species that were produced by a plasma discharge in the presence of

steam (Eq. 5.20) had higher densities due to the external oscillation [131, 135].

The destruction energy efficiency had a pattern that was similar to that of the decomposition

efficiency. With the increase in the steam feed rate, the destruction energy efficiency

Oscillation amplitude (Vpp

)

De

co

mp

ositio

ne

ffic

ien

cy

(%)

De

str

uctio

ne

ne

rgy

eff

icie

ncy

(g/k

Wh

)

Be

nze

ne

co

nce

ntr

atio

n(%

)

0.5 1 1.5 2 2.5 30

20

40

60

80

100

0

5

10

15

20

25

30

0

0.05

0.1

0.15

0.2

0.25

0.3




Input concnetration

(a)

H2,C

O,C

O2

(%)

Ca

rbo

nb

lack

(g/N

m3)

0

0.5

1

1.5

2

0

0.02

0.04

0.06

0.08

0.1

Oscillation amplitude (Vpp

)

CH

4,C

2H

4,C

2H

6(%

)

0.5 1 1.5 2 2.5 30

0.05

0.1

0.15

0.2

C2H

6

H2

CO

Carbon black

CH4

C2H

4

CO2

(b)

97

increased and reached 22.9 g/kWh at the steam feed rate of 0.66 L/min. Then it gradually

decreased. As shown in Eq. 5.2, the destruction energy efficiency increased due to the

increase in the benzene tar removal and the input feed rate with the increase in the steam feed,

while supplied to a specific quantity of the input electric power. After the maximum was

reached, however, the destruction energy efficiency decreased because of the effect of the

decreased benzene tar removal, despite the increase in the steam feed rate.

The amount of carbon-black was relatively large at 0.053 g/Nm3, when no steam was fed (0

L/min). This was because the benzene tar was decomposed into carbon (C) due to the carbon

formation (Eq. 5.6) without the oxidation caused by the OH radicals. As the steam feed rate

increased, carbon-black gradually decreased and was hardly produced when the steam feed

rate was 0.38 L/min or higher. This was because the produced carbon-black was converted

into light gases according to the soot decomposition (Eqs. 5.10 and 5.11), due to the OH

radicals that were produced according to the water excitation (Eq. 5.7) [26].

Figure 5.26(b) shows the concentration of light gases, which were produced when the

benzene tar was decomposed.

With the increase in the steam feed rate, H2 and CO2 continued to increase, and CO increased

to the maximum and then decreased. H2, CO2, and CO mostly increased due to the tar

cracking (Eq. 5.5), the carbon formation (Eq. 5.6), the soot decomposition (Eqs. 5.10 and

5.11), and the steam reforming (Eq. 5.13). In the case of CO, however, it decreased later as it

was converted into CO2 due to the radical termination (Eq. 5.9) and the water-gas shift

reaction (Eq. 5.12).

Figure 5.26 Effects of the steam feed rate

When no steam was fed, light hydrocarbon gases (CH4, C2H4, and C2H6) were produced with

the concentrations of 0.1, 0.09, and 0.11%, respectively. They decreased with the increase in

the steam feed rate and were hardly produced at the steam feed rate of 0.47 L/min or higher.

The hydrocarbons were produced as the benzene tar was decomposed due to the tar cracking

(Eq. 5.5), and disappeared due to the steam reforming (Eq. 5.13) as steam was fed.


De

co

mp

ositio

ne

ffic

ien

cy

(%)

De

str

uctio

ne

ne

rgy

eff

icie

ncy

(g/k

Wh

)

Be

nze

ne

co

nce

ntr

atio

n(%

)

0 0.2 0.4 0.6 0.80

20

40

60

80

100

0

5

10

15

20

25

30

0

0.05

0.1

0.15

0.2

0.25

0.3


Decomposition efficiency (With oscillation)


Input concnetration

(a)

Decomposition efficiency (Without oscillation)

H2,C

O,C

O2

(%)

Ca

rbo

nb

lack

(g/N

m3)

0

0.5

1

1.5

2

0

0.02

0.04

0.06

0.08

0.1


CH

4,C

2H

4,C

2H

6(%

)

0 0.2 0.4 0.6 0.80

0.05

0.1

0.15

0.2

C2H

6

H2

CO

Carbon black

CH4

C2H

4

CO2

(b)

98


Figure 5.27 shows the effects of the change in the total gas feed rate. The total gas feed rate

was set within the 12.6~24.7 L/min range, which can be stably operated for the EOPR.


concentrations at the inlet and outlet of the EOPR, with and without oscillation.

With the increase in the total gas feed rate, the decomposition efficiency slightly decreased.

This was because the gas feed rate increased in the EOPR with the increase in the total gas

feed rate, and the retention time of the benzene-containing gas decreased within the EOPR.

Therefore, the reaction time between the electrons, ions, and radicals that were produced in

the plasma discharge zone, and the benzene decreased [106].

Without oscillation, the decomposition efficiency for the change in the total gas feed rate was

almost the same pattern as that with external oscillation. The decomposition efficiency,

however, was higher by 8.91% with the external oscillation. This indicates that the external

oscillation effect is almost constant regardless of the total gas feed rate.

The destruction energy efficiency significantly increased with the increase in the total gas

feed rate. This is because the gas feed rate (Q) relatively increased, although the benzene

removal slightly decreased according to the increase of steam feed.

Figure 5.27(b) shows the concentrations of carbon-black and light gases, which were

produced when the tar benzene was decomposed.

Carbon-black was hardly generated with the change in the total gas feed rate. This was

because even though the benzene concentration slightly increased with the increase in the

total gas feed rate, carbon-black was destructed into light gases (Eqs. 5.10 and 5.11) at a

constant and sufficient steam flow rate of 0.66 L/min.

Figure 5.27 Effects of the total gas feed rate

With the increase in the total gas feed rate, H2 and CO gradually increased. H2 increased due

to the tar cracking (Eq. 5.5) and the carbon formation (Eq. 5.6), and H2 and CO increased due

to the soot decomposition (Eq. 5.10) and the steam reforming (Eq. 5.13), when steam existed.

The increase in CO2 was less than those in the two gases above. This was because the reaction

did not last sufficiently as the steam feed rate was not high enough due to the radical

termination (Eq. 5.9), the soot decomposition (Eq. 5.11), and the water-gas shift reaction (Eq.


De

co

mp

ositio

ne

ffic

ien

cy

(%)

De

str

uctio

ne

ne

rgy

eff

icie

ncy

(g/k

Wh

)

Be

nze

ne

co

nce

ntr

atio

n(%

)

12 14 16 18 20 22 240

20

40

60

80

100

0

10

20

30

40

50

0

0.05

0.1

0.15

0.2

0.25

0.3


Decomposition efficiency (Without oscillation)


Input concnetration

(a)

Decomposition efficiency (With oscillation)

H2,C

O,C

O2

(%)

Ca

rbo

nb

lack

(g/N

m3)

0

0.5

1

1.5

2

0

0.02

0.04

0.06

0.08

0.1


CH

4,C

2H

4,C

2H

6(%

)

12 14 16 18 20 22 240

0.05

0.1

0.15

0.2

C2H

6

H2

CO

Carbon black

CH4

C2H

4

CO2

(b)

99

5.12). Light hydrocarbon gases (CH4, C2H4, and C2H6) were hardly detected. This was

because the hydrocarbons were converted into H2 and CO due to the steam reforming (Eq.

5.13).

Eventually, the increase in the total gas feed rate reduced the retention time, and then the

decomposition efficiency decreased, but because of the characteristics of the test equipment,

the input benzene concentration slightly increased, and the light gases other than hydrocarbon

also increased.

5.4 Summary

The GAPR which developed in the biogas reforming study (Chapter 4) was tested for

verification of tar destruction. The selected surrogate tar was benzene for light aromatic

representative tar and anthracene for light PAH representative tar. And the GAPR was also

tested for the verification of real tar destruction combined with a continuous screw pyrolyzer.

In addition, the EOPR was designed and the test was conducted to enhance the idea of the

plasma reformer.

▌For light aromatic tar result, a parametric screening study was achieved to show the

optimal operating conditions and the best design guide line. The operating parameters were

the steam feed rate, the input benzene concentration, the total gas feed rate, and the input

electric energy. And the design factors changing were the nozzle diameter, the electrode gap,

the electrode length, and the electrode shape.

The optimal operation conditions were the steam feed rate of 0.66 L/min, the benzene

concentration of 0.12%, the total gas feed rate of 16.7 L/min, the SEI of 0.17 kWh/m3, the

nozzle diameter of 3 mm, the electrode gap of 3 mm, the electrode length of 95 mm, and the

Arc 1 electrode shape. The maximum decomposition efficiency was 82.6%, and the

destruction energy efficiency was 20.9 g/kWh. For the optimum design, the ratio of the

electrode gap to the nozzle diameter must be 1 or higher. In addition, the electrode type of the

Arc 1 showed the highest decomposition and destruction energy efficiency because it ensured

a sufficient plasma discharge column.

▌For light PAH tar result, the steam feed rate, the input tar concentration, the total gas feed

rate, and the input electric energy change were used as variables for the anthracene test. When

steam was fed at the rate of 0.63 L/min, the decomposition efficiency was highest (96.1%)

due to the creation of the OH radicals. The destruction energy efficiency was highest (2.63

g/kWh) when the total gas feed rate was 24.1 L/min. H2, CO, and CO2 were produced as

reforming gases. At the steam feed rate of 0.37 L/min or more, carbon-black did not appear.

The higher heating value of the gas produced from the anthracene decomposition was 11,324

kJ/Nm3, and the carbon balance was 87.6%.

▌For light real tar result, the gravimetric tar mass was significantly reduced to 3.74 g/Nm3 at

the outlet of the plasma reformer from 18.02 g/Nm3 at the inlet of the plasma reformer. The

destruction efficiency was 79.2%, accordingly. The concentrations of the light tar compounds

were also significantly reduced to 0.46 g/Nm3 from 3.47 g/Nm

3 for benzene, 0.11 g/Nm

3 from

0.37 g/Nm3 for naphthalene, 0.03 g/Nm

3 from 0.09 g/Nm


3 from

0.07 g/Nm3 for pyrene, 0.06 g/Nm

3 from 0.85 g/Nm

3 for benzonitrile, and 0.0 g/Nm

3 from

0.04 g/Nm3 for benzoacetonitril, after the reforming process.

100

▌As for the EOPR test, the oscillation frequency, the oscillation amplitude, the steam feed

rate, and the total gas feed rate were used as parameters for the test. The optimum oscillation

frequency, the oscillation amplitude, and the steam feed rate conditions existed for the

maximzing the decomposition and energy efficiencies. That is, when the oscillation frequency

was 267 Hz, the oscillation amplitude was 3 Vpp, and the steam feed rate was 0.66 L/min, the

destruction and energy efficiencies were 90.7% and 22.95 g/kWh, respectively. With the

increase in the total gas feed rate, the decomposition efficiency slightly decreased, and the

destruction energy efficiency increased. The benzene tar was decomposed into light gases (H2,

CO, and CO2), hydrocarbons (CH4, C2H4, and C2H6), and carbon-block.

Without oscillation, the decomposition efficiency was 82.6%, and the destruction energy

efficiency was 20.9%, both of which were 8.9% lower than those with the external oscillation.

The test results showed that the EOPR can efficiently destruct benzene tar using less energy

than that used in the gliding arc plasma reformer.

101

Chapter 6

Sequential carbonization-activation system including

char production and tar removal

For the conversion of sewage sludge to energy and resource utilization, a novel sequential in-

line thermal treatment system was suggested. A research approach was achieved in 2 steps in

a pilot scale test rig for practical system design.

First, a combined carbonization-activator was developed to improve tar adsorption capability

of the sludge char and to achieve high producer gas yield. To determine the optimal operating

conditions, a parametric study was conducted on the steam feed rate, the activator temperature

and the sludge moisture content.

Second, an integrated sludge treatment system with in-line connection of the carbonization-

activator, the plasma reformer, and the fixed bed adsorber was suggested and verified for its

performance. The carbonization-activator produced sludge char and producer gas containing

tar. The plasma reformer was set to improve the producer gas yield by destructing tar released

from the carbonization-activator. The fixed bed adsorber, filled with the sludge char produced

from the carbonization-activator, was installed for adsorption of un-treated residual tar.

In addition, the process analysis for the sequential in-line carbonization-activation system was

conducted. The calculations of the mass and energy balances for each component (combined

carbonization-activator; gliding arc plasma reformer; fixed bed adsorber) were conducted.

The sequential carbonization-activation system was performed for total system efficiency.

Further, for the improvement of the total system efficiency, two scenarios were suggested,

and the verification of the system efficiency improvement was performed.


Most of waste sludge was treated through landfill, incineration, and land spreading [4, 62, 69].

However, landfill requires the complete isolation between filling site and surrounding area

due to leaching of hazardous substance in sludge, and has the limited space for filling site.

Utilization of sludge as compost incurs soil contamination by increasing the content of heavy

metal in soil, and causes air pollution problem due to dispersing of hazardous component to

atmosphere. Incineration has the benefits of effective volume reduction of waste sludge and

energy recovery, but insufficient mixing of air could discharge hazardous organic pollutant

especially in the condition of low oxygen region. In addition, significant amount of ashes with

hazardous component will be created after incineration.

As alternative technology for the previously described sludge treatment methods, researches

on pyrolysis [4, 72, 135] and gasification treatment [136, 137] have been conducted.

Pyrolysis or gasification can produces gas, oil, and char that could be utilized as fuel, adsorber

and feedstock for petrochemicals. In addition, heavy metal in sludge (excluding cadmium and

mercury) can be safely enclosed. It is treated at the lower temperature than incineration so that

amount of contaminant is lower in producer gas from pyrolysis or gasification processes due

to no or less usage of air. Moreover, hazardous components, such as dioxin, are not generated.

102

However utilization of the producer gas into an engine, a gas turbine, fuel cell, etc might

cause the condensation of tar. In addition, aerosol and polymerization reaction could cause

clogging of cooler, filter element, engine inlet, etc [26, 138].

As the removal methods of tar component, In-Furnace Technology (IFT) and After Treatment

Technology (ATT) were suggested.

Firstly, the IFT does not require the additional post-treatment facility for tar removal, and

further development is required for optimizing the operating condition and developing novel

designs of the pyrolyzer or gasifier. Through these conditions and technical advancement,

production of syngas with low tar content can be achievable, but cost and large scaled

complex equipments are needed [11, 139].

Secondly, multi-faceted researches on the ATT, such as the thermal cracking [81, 140], the

catalysis [141], the adsorption [81], the steam reforming [81, 142, 143], the partial oxidation

[81, 143], the plasma discharge [26, 109, 123, 130, 144-146], etc have been conducted. For

the thermal cracking, higher than 800℃ is required for the destruction, and its energy

consumption surpass the production benefit. The catalyst may reacts with contaminants such

as sulfur, chlorine, nitrogen compounds from biomass gasification. Also, the catalyst can be

de-activated due to coke formation, and additional energy cost to maintain high temperature is

needed. For the adsorption, there were several researches utilizing char, commercial activated

carbon, wood chip and synthetic porous cordierite for tar adsorption [72, 81]. In case of

adsorbers having mesopore, the adsorption performance of light PAH tars, such as

naphthalene, anthracene, pyrene, etc excluding light aromatic hydrocarbon tar (benzene,

toluene, etc) was superior [81]. Tar reduction in the steam reforming, the partial oxidation and

the plasma discharge can produce syngas having major compounds of hydrogen and carbon

monoxide through the reforming and cracking reaction. The steam reforming has a good

characteristic in high hydrogen yield. But it requires high temperature steam which consumes

great deal of energy. In addition, longer holding time might require for larger facility scale.

On the contrary, the partial oxidation reforming features less energy consumption, and has the

benefit of heat recovery due to exothermic reaction. However, the hydrogen yield is relatively

small, and a large amount of carbon dioxide discharge is the disadvantage. Researches on tar

decomposition via the plasma discharge were conducted in dielectric barrier discharge (DBD)

[130], the single phase DC gliding arc plasma [26, 123, 146] and the pulsed plasma discharge

[144]. Compared to the conventional thermal and catalytic cracking, the plasma discharge

shows the higher removal efficiency due to the formation of radicals. However, high cost of

preparation of power supply and short life cycle is the key for improvement. A 3-phase arc

plasma applied for tar removal is easy to control the reaction, and has high decomposition

efficiency along with high energy efficiency.

That is to say; all the methods have limitation in the waste sludge treatment for producing

products and removing tar in the producer gas. Therefore, the combination of both IFT and

ATT should be accepted as a new alternative method for improving the environment-

friendliness.

In this study, a combined carbonzation-activator was designed for the sequential pyrolysis and

steam activation processes. The combined carbonzation-activator was composed of a screw

carbonizer and a rotary drum activator for production of high porosity sludge char and clean

producer gas, simultaneously. Parametric researches were conducted on the steam feed rate,

the activator temperature and sludge moisture content to obtain the optimal operating

conditions and design characteristics of each variable.

103

Then a sequential sludge thermal treatment system was suggested for the production of high

quality gas and sludge char from waste sewage sludge. The system is composed a

carbonization-activator, a gliding arc plasma reformer, and a fixed bed adsorber. The system

analysis on the carbonization-activation characteristics and tar reduction from the thermal

treatment system was achieved.

In addition, the process analysis for the sequential carbonization-activation system was

performed for verifying the system efficiency, including the mass and energy balance analysis

for each component of the system; the carbonization-activator, the plasma reformer, and the

adsorber.


6.2.1 Dried sludge for experiment

Sewage sludge from a local waste water treatment plant was dewatered by centrifuging. The

moisture content of the dewatered sludge was about 80%. For the carbonization and activation,

high moisture content in the dewatered sludge will give significant energy loss due to

preemptive utilization of the heat for drying. In addition, the moisture will affect the product

gas due to its reaction with other reactants at the stage of delayed pyrolysis and gasification.

Therefore, the dewatered sludge was dried to less than 10% moisture content using the rotary

drum dryer developed in this study (Refer Chapter 2). The properties of the dried sludge are

shown in Table 6.1. The dried sludge for use in the experiments was sieved to homogeneous

size distribution of 5~10 mm, using a Taylor sieve (Ro-Tap Sieve Shaker, Chunggye Ltd.,

Korea).

Table 6.1 Properties of the dried sludge

Proximate analysis (%)

Moisture Volatile matter Fixed carbon Ash

9.7 51.7 6.1 32.5

Ultimate analysis (%)

C H O N S

52.3 8.2 32.2 7.92 0.01

6.2.2. Carbonization-activation experiment

The dewatered sludge from a waste water treatment plant can be treated in a sequential in-line

treatment system as shown Figure 6.1. The rotary drum dryer in the figure was separately

used as shown Chapter 2. For this Chapter 6, a carbonization-activation system was composed

of the pyrolysis gasifier, the 3-phase gliding arc plasma reformer, and the fixed bed adsorber,

as shown in Figure 6.2.

The carbonization-activator was designed to be a combined rig with the screw carbonizer for

pyrolysis of dried sludge and the rotary activator for steam activation of carbonized material.

The screw carbonizer was manufactured as a feed screw type for carbonization of dried

sludge. The feed screw controls the holding time of the dried sludge in the carbonizer

according to motor revolution number. The screw carbonizer consists of co-axial dual pipes.

104

The superheated steam is fed to a gap between inner and outer pipes, and then radially injects

to the rotary activator. This configuration, newly designed in this study, should be superior in

view of the energy saving. That is, the dewatered sludge in the screw carbonizer can receive

the steam heat by convective and conductive heat transfers without surface heat loss. That is

why the carbonizer and activator were designed as the combined configuration. The rotary

activator is composed of the rotary drum with a vane and pick-up flight, the indirect heating

jacket, the gas sampling port, the char outlet, etc. The retention time of activating sludge is

controlled by the rotation speed of the rotary drum. The sludge feeding device is designed for

holding the dried sludge in a dried sludge hopper which is installed at the inlet of the

combined carbonization-activator. The screw feeder is installed at the bottom of the hopper,

and controls the input amount of dried sludge by the revolution speed. The feeder feeds the

dried sludge into the screw carbonizer.

Figure 6.1 Experimental setup of a sequential in-line treatment system

The hot gas generator is designed for producing a hot combustion gas to heat the heating

jacket and supplies hot steam into the rotary drum. It was composed of a combustor with a

burner and a steam generator.

The 3-phase gliding arc plasma reformer was installed at downstream of the outlet of the

carbonization-activator. The gliding arc plasma reformer utilized a quartz tube (55 mm in

diameter, 200 mm in height) for insulation and monitoring plasma discharge state. The three

fan-shaped electrodes (95 mm in length) were installed with the interval angle of 120 degree

on a ceramic connector (Al2O3, 96 wt.%) for complete insulation between the electrodes. The

gap of each electrode was 3 mm. At the inlet of the plasma reformer, an orifice disc with 3

mm hole for the injection of producer gas was installed. A 3-phase AC high voltage power

supply unit (Unicon Tech., UAP-15K1A, Korea) was used for stable plasma discharge inside

of the plasma reformer.

The fixed bed adsorber was installed at downstream of the plasma reformer. The adsorber was

made of a stainless steel cylinder. The capacity of the adsorption tower was 730 mL (76 mm

in diameter, 160 mm in length). A straight honeycomb ceramic distributer was installed inside

105

of the cylinder. The sludge char was put on a wire-mesh plate which poisoned on the top of

the distributer.

All the tests were conducted by checking the stabilization status in the carbonization-activator,

while measuring the temperatures of the screw carbonizer, the rotary activator and the hot gas

generator, etc.

Figure 6.2 Detailed front view of the carbonization-activation system

Figure 6.3 presents the temperatures in selected components under the optimal operating

conditions as an illustration. While the hot gas generator provided combustion gas to the

heating jacket for heating the carbonization-activator, the temperatures of both the carbonizer

and the activator were gradually increased. When the activator temperature was raised above

that of the steam activation, the dried sludge was introduced. And after the temperature of the

carbonizer was stably maintained at about 500℃ and the activator temperature maintained at

820℃, sludge char and producer gas samples were collected from the char outlet and the

sampling port, respectively.

Figure 6.3 Initial operating and stabilization characteristics of the carbonization-activator

Elapsed time (hr)

Te

mp

era

ture

(oC

)

0 1 2 3 4 5 6 7 8 9 100

200

400

600

800

1000

Starting test pointfor samplingsSteam feeding point

1 Combustor

Dried sludge feeding point

2 Rotary activator

3 Screw carbonizer

Steam generator4

106

▌For combined carbonization-activator test, an experiment was conducted to assess for the

steam feed rate, the activator temperature and the sludge moisture content, which were major

factors for the performances of the adsorption capability of the sludge char and formation of

producer gas. Table 6.2 describes experimental conditions for the parametric study.

Through the parametric study, the optimal operating conditions (Table 6.3) were taken as the

state that gave high quality and large amounts of the sludge char and producer gas. The

optimal conditions were used for three cases to evaluate the tar adsorption performance of the

sludge char, the mass and heat balances, and the performance of the carbonization-activator

system.

Table 6.2 Experimental conditions for each variable

Condition Steam feed rate (mL/min) Activator temperature (℃) Sludge moisture content (%)

Variable

range 0 ~ 30 650 ~ 820 9 ~ 18

▌For carbonization-activator system test, experiments were conducted at the optimal

condition for high quality porosity in the sludge char and for the largest amount of clean

producer gas. The experimental conditions including the operating temperatures are given in

Table 6.3. All the data in experiments were taken after stabilizing the temperatures in each

component, particularly in the screw carbonizer and the rotary activator. After finishing

experiment, the sludge char in the char outlet was cooled down to a room temperature by

passing nitrogen through the carbonization-activator to protect the oxidation of the sludge

char by air. The producer gas was sampled for 5 min in a stainless cylinder at the sampling

ports of the pyrolysis gasifier, the plasma reformer, and the adsorber. Tar sampling was

conducted for 20 min according to the sampling method, and the total amount of gas was

measured with a gas-flow meter.

In addition, the mass and energy balance analysis in the combined carbonization-activator

system were achieved at the optimal condition (Table 6.3).

Table 6.3 The optimal conditions and operating temperatures in each component

Test operating conditions

Steam feed rate

(mL/min) Sludge moisture content (%)

1)

Retention time (min)

Activator Carbonizer

10 11 30 30

Temperature (℃) in each component

① Combustor ② Carbonizer ③ Activator ④ Steam generator ⑤ Plasma

reformer ⑥ Adsorber

1,010 450 820 450 400 35

<Note> 1)

Moisture content of dried sludge is average number

In each test, the gas and tar sampling were conducted 2 or 3 times in the test duration at a

stable condition, and the data were averaged to take the probability of the analysis data. The

107

measurement for tar, producer gas, and sludge char was conducted following the procedure

stated in “3.2.3 Sampling and analysis method for products”.


6.3.1 Combined carbonization-activator

Parametric researches were conducted on the steam feed rate, the activator temperature and

the sludge moisture content to investigate the design characteristics of each variable and to

obtain the optimal operating conditions.

▌Effects of the steam feed rate

Figure 6.4 shows the producer gas concentration and its higher heating value by changing the

steam feed rate in the range of 0~30 mL/min. The moisture content of dried sludge and the

temperature of the carbonization-activator were fixed at 10% and 820℃, respectively.

When the steam feed rate was increased, H2 concentrations increased by the water gas

reaction (Eq. 3.2) and tar steam gasification (Eq. 3.6). And the produced CO decreased and

CO2 increased by the water-gas shift reaction (Eq. 3.8). The CH4 contents decreased due to

the methanation inverse reaction (Eq. 3.7) at a high temperature. Similar tendency is shown

by Umeki’s research [87].

The higher heating value of the producer gas decreased with the increasing of the steam feed

rate. This is because the combustible gases, such as CO and CH4, decreased, while the non-

combustible gas, CO2, increased. And an increase of H2 had a minor effect due to the lower

heating value compared to CH4. Also, CO2 diluted the producer gas resulting in the decrease

of the heating value.


Figure 6.5 shows the SEM photographs (in 2,200 times of magnification) of the pore

development and distribution in sludge char by changing the steam feed.

Steam feed rate (mL/min)

Ga

sco

nce

ntr

atio

ns

(%)

Hig

he

rh

ea

tin

gva

lue

(kJ/N

m3)

0 5 10 15 20 25 300

10

20

30

40

50

0

2000

4000

6000

8000

10000

12000

CH4

CO

CO2

H2

HHV

108

Figure 6.5(a) shows the carbonization only without steam feed. Figure 6.5(b)~(d) represent

the cases with the stream feed of 10, 20 and 30 mL/min, respectively.

The fed steam penetrates the internal pores of the carbonized char, with micro-pores

developed via the oxidization (i.e., the water gas reaction; Eq. 3.2) with surface organic

compounds. Different pore distributions and morphologies can be acquired by varying the

steam feed rate.

When steam was not fed, as shown in Figure 6.5(a), the micro-pores formed were minimal

and small. However, the stream fed at 10 mL/min, as shown in Figure 6.5(b), allowed the

development of a largest proportion of micro-pores compared to the other steam feeds. With

the steam fed at 20 mL/min or more as shown in Figure 6.5(c) and (d), the micro-pores on the

surface of the activated char deteriorated due to a change in the surface state from rough to

smooth due to the adhesion of the high temperature steam. This is why the size and

distribution of the micro-pores were reduced with higher steam feed.

To evaluate the adsorption characteristics of the sludge chars produced under different steam

input conditions during the sequential carbonization and activation, a fixed bed adsorption

tower was used. The input concentration of benzene in the adsorption test was fixed at 5,200

ppm, with the H/D ratio (i.e., ratio of tower height and diameter) of 2 and the GHSV (gas

hourly space velocity) of 1,175/h at 20℃.

Figure 6.5 SEM photographs of the sludge chars with different steam feeds

Figure 6.6 shows the breakthrough curve, the amount of adsorption and saturation point for

comparing the adsorption characteristics of the activated sludge char. In Figure 6.6, C is the

effluent concentration and Ci is the input concentration; the SFR (steam feed rate) designates

the amount of the input steam.

With the SFR of 10 mL/min, the saturation point of the sludge char was longest at 45 min,

and the amounts adsorbed showed the largest value of 140 mg/g. The time to reach the

saturation point with SFR of 0, 20 and 30 mL/min were 25, 20 and 17 min, respectively. The

amounts of adsorption accordingly decreased to 112, 97, and 84 mg/g, respectively. This

trend can be explained because with SFR of 10 mL/min the largest amount of micro-pores

developed.

109

Figure 6.6 Adsorption characteristics of benzene onto sludge chars with different steam inputs

▌Effects of the temperature of the carbonization-activator

Figure 6.7 presents the analysis of the pyrolysis gas with changes in the temperature of the

carbonization-activator from 650 to 820℃. The moisture content of the dried sludge was set

at about 10%, and the amount of steam input was fixed at 10 mL/min.

With increasing the temperature of carbonization-activator, the amounts of H2 and CO

increased due to the thermal cracking reaction and the steam reforming of volatile organic

compounds originating from the decomposition of the organic matter contained in the sewage

sludge. However, the amounts of CH4 and CO2 increased and then slightly decreased after

showing their maximum concentrations. At the temperature higher than 750℃, CH4 and CO2

were decomposed into H2 and CO due to the reverse reactions of the methanation reaction (Eq.

3.7) and the water gas shift reaction (Eq. 3.8)

Figure 6.7 Effect of the activator temperature

Steam feed rate (mL/min)

Ad

so

rptio

na

mo

un

t(m

g/g

)

Sa

tura

tio

np

oin

t(m

in)

0 10 20 30 400

20

40

60

80

100

120

140

160

180

200

40

20

0

20

40

60

Adsorption time(min)

C/C

i

0 10 20 30 40 500

0.2

0.4

0.6

0.8

1

SFR=10 mL/min

SFR=0 mL/min

SFR=20 mL/min

SFR=30 mL/min

Temperature (oC)

Ga

sco

nce

ntr

atio

ns

(%)

Hig

he

rh

ea

tin

gva

lue

(kJ/N

m3)

650 700 750 8000

10

20

30

40

50

0

2000

4000

6000

8000

10000

12000

H2

CO

CH4

CO2

HHV

820

110

The heating value of the producer gas increased with increasing the carbonization-activator

temperature due to the increased formations of H2 and CO from CH4. In addition, the

decomposition of organic matter inside of the dried sludge was also enhanced.

Figure 6.8 show the SEM photographs of the surfaces of the activated sludge char at different

carbonization-activator temperatures. The pore on the surface of the sludge char displayed

reduced smooth faces with increasing the temperature from 650 to 800℃, and the pore

development was then well established. The SEM in Figure 6.8(e) presents the best pore

development at 820℃, which was found to be the optimal operating condition for this

carbonization-activator.

Figure 6.8 SEM photographs of the sludge char at different carbonization-activator

temperatures

▌Effect of the sludge moisture content

Figure 6.9 presents the concentrations of the producer gas with changes in the moisture

content of the dried sludge. Moisture content of the dried sludge varied from 9 to 18%, while

the carbonization-activator temperature and the amount of steam feed were fixed at 820℃

and 10 mL/min, respectively.

With increasing the moisture level in the sludge, the concentrations of H2, CH4, and CO

decreased because of the reduced amounts of producer gas from the slow pyrolysis rate with

increasing the moisture content in the dried sludge.

However, the CO2 concentration was increased, which can be explained by the water-gas shift

reaction (Eq. 3.8) and due to the increased moisture level.

The heating value of the producer gas decreased with increasing the moisture level in the

dried sludge, which could be explained by the reduced concentrations of flammable gases (i.e.,

H2, CH4, and CO) and the dilution effect of the CO2.

111

Figure 6.9 Effect of the sludge moisture content

Figure 6.10 show SEM photographs of the surface of the sludge char with changes in the

moisture level in the dried sludge.

The moisture content of the dried sludge affected the pore development because they are

formed by the evaporation of the moisture from the dried sludge. The specific moisture

content, with an optimal condition shown in Figure 6.10(b), displayed the largest

development of micropores. It is because the moderate amount of moisture in the sludge char

promotes the development of micropores, which is desirable for high adsorptive capacity.

That is, this gives oxidation thermosetting of the sludge surface due to oxidants (H2O or O2)-

carbon interactions thus inhibiting the formation of softened intermediate products during the

carbonization-activation process [63]. However, at higher than 11% moisture level in the

sludge char, less formation of micro-pore should be observed by the lower development of

pores due to delayed water gas reaction in the pore surface.

Figure 6.10 SEM photographs with respect to the moisture level in the dried sludge

Moisture content (%)

Ga

sco

nce

ntr

atio

ns

(%)

Hig

he

rh

ea

tin

gva

lue

(kJ/N

m3)

9 11 13 15 170

10

20

30

40

50

0

2000

4000

6000

8000

10000

12000

CO2

CO

CH4

HHV

H2

18

112

6.3.2 Sequential in-line carbonization-activation system

1) Characteristics of the combined carbonization-activator

Figure 6.11 shows the mass yield of the sludge char, the tar, and the producer gas from the

combined carbonization-activator. The product amounts of the sludge char, the tar and the

producer gas were 35.4%, 21% and 43.6%, respectively. As described before, the experiment

setup was made to primary pyrolysis carbonization at the screw carbonizer which was set at

450℃ and the post steam activation at the rotary activator which was set at 820℃.

The producer gas was formed by decomposition and volatilization of organic compound in

the screw carbonizer, and gas formation was increased due to the steam reforming of the tar

and the sludge char in the rotary activator. The sludge char yield was reduced by vaporization

of the volatile component during passing of the screw carbonizer, and by the steam

gasification and inorganic decomposition in the rotary activator. Heavy tar formed was

converted into the producer gas and light tar in the rotary activator.

Figure 6.11 Mass yield of the products

▌Characteristics of the sludge char

Figure 6.12 compares incremental pore volume and SEM photos of the dried sludge and the

sludge char. The pore size classification in this study follows the IUPAC classification [63,

147] i.e. micropores (<2 nm), mesopores (2~100 nm) and macropores (>100 nm). Pore of the

sludge char after the carbonization-activation showed significant increase compared to the

dried sludge, and the pore distribution was mostly less than 50 nm, which is comprised of

micropores and mesopores.

The carbonization-activator was designed as a continuously combined type for carbonization

of dried sludge in the screw carbonizer and the steam activation in the rotary activator. The

dried sludge experienced evaporating of moisture and decomposing of organic component for

pore development through passing the screw carbonizer [147]. And then the carbonized

material was exposed to steam in the rotary activator for the formation and development of

micropoees and mesopores. For the steam activation in developing micropores, steam should

Ma

ss

yie

ld(%

)

0

10

20

30

40

50

GasChar Tar

43.6

35.4

21.0

113

deeply penetrate into pores of the carbonized material for the surface reaction. High

temperature activation had the benefit of diffusion and penetration of the steam to develop

micropore. On the other hand, the steam was blocked by tar in the carbonized material,

resulted in well-developed mesopore due to its reaction on the surface. This is the reason why

both micropores and mesopores were developed in the sludge char from the carbonization-

activation.

Sludge drying was conducted with the parallel flow rotary drum drier (refer to Chapter 2)

with direct-hot gas application. Hot gas inflow in turbulent flow directly contacted with the

dewatered sludge in the rotary drum dryer. Inside of the dryer, the temperature was set at

255℃ in average value. For the dried sludge, a small portion of micropore and mesopore was

formed at the dryer temperature. It is considered to be formed due to discharging of volatile

organic material and dehydroxlation of inorganic material from the dried sludge.

Bagreev et al. proved that water released by the dehydroxylation of inorganic material could

aid pore formation and moreover could act as an agent for creating micropores [93]. In

addition, Inguanzo et al. proposed that carbonization increases the porosity through

unblocking of many of the pores obscured by volatile matter [94].

Surface of the dried sludge from shown in SEM photograph in 50,000 times of magnification

presents smooth surface with less pores, but the sludge char presents overall formation of

pores.

Figure 6.12 Incremental pore volume and SEM images of the dried sludge and sludge char

The characteristics of the sludge char was measured and the mean pore size, the specific

surface area and the pore volume were 6.35 mm, 98.1 m2/g, 0.2354 cm

3/g, respectively. The

mean pore size of the sludge char is mesopore. It might be expected that the sludge char have

similar characteristics with results of Phuphuakrat et al for tar adsorption, showing

particularly good adsorbability in condensable light PAHs (e.g. naphthalene, anthracene,

pyrene) [81].

A semi quantitative chemical analysis of the dried sludge and the sludge char, shown in

Figure 6.13 and Table 6.4, was obtained from the EDX analyzer coupled to SEM

measurements. The results indicate that both samples present relatively high carbon content in

addition to mineral components. The relative amount of carbon decreased after carbonization

and activation, as expected considering the decomposition of the organic components.

Pore width (nm)

Incre

me

nta

lp

ore

vo

lum

e(c

m3/g

)

0

0.0005

0.001

0.0015

0.002

0.0025

0.003

1002 50

Sludge char

10

Dried sludge

2 nm2 nm

114

These atoms might be considered as potential catalysts for the pyrolysis reaction. For example,

with Al, if existing in the form of Al2O3, it would be an acid catalyst for the cracking reaction

[96]; or with K, and Ca atoms, which have already been reported as the catalyst for biomass

pyrolysis in literature [97].

Figure 6.13 EDX spectrums of the dried sludge and the sludge char

Table 6.4 Elements content of the dried sludge and sludge char

Item C O Mg Al Si P S Cl K Ca Ti Fe Zn Ba

Dried sludge

(wt.%) 53.65 44.62 0.06 0.23 0.45 0.55 0.03 0.01 0.06 0.07 0.01 0.24 0.02 0

Sludge char

(wt.%) 47.65 44.83 0.14 1.21 5.34 0.46 0.03 0.02 0.09 0.11 0 0.21 0 0.01

Figure 6.14 shows the N2 adsorption-desorption isotherm for the dried sludge and the sludge

char. According to the isothermal adsorption graphs, the dried sludge exhibited only a small

amount of adsorption, but the sludge char displayed a larger amount of adsorption at lower

nitrogen concentrations. As shown in Figure 6.12, the sludge char exhibited well-developed

micro- and meso-pore structures. The analysis on the adsorption isotherm provides an

assessment for the pore size distribution. According to the IUPAC classification, the curve of

the sludge char corresponds to the Type V isotherm. A characteristic of the Type V isotherm

is the hysteresis loop, which is associated with the capillary condensation in mesopores and

limiting uptake at a relatively high pressure [95].

Figure 6.14 Isothermal adsorption-desorption linear plot

Energy (keV)

Co

un

ts

2 4 6 8 100

200

400

600

800

1000

C

Sludge char

Mg

Zn

Al

Si

P

SCl K Ca

Ti Ba Fe

O

Dried sludge

0


Ad

so

rbe

da

mo

un

to

fN

2(c

m3/g

)

0 0.2 0.4 0.6 0.8 10

20

40

60

80

100

120

140

160

180

Sludge char desorptionSludge char adsorption

Dried sludge adsorption

Dried sludge desorption

115

▌Characteristics of the tar

Gravimetric tar and selected lights tar produced from the carbonization-activator were shown

in Table 6.5. The light tars for the corresponding benzene ring were selected to benzene (1

ring), naphthalene (2 ring), anthracene (3 ring) and pyrene (4 ring). And the light tars

generated from nitrogen component in the sewage sludge [4] were taken as benzonitrile and

benzeneacetonitrile.

Gravimetric tar mass was 26.3 g/Nm3. The total concentration of light tar was 10.9 g/Nm

3,

and its amount order was benzene, naphthalene, benzonitrile, benzeneacetonitrile, anthracene,

and pyrene.

Dried sludge formed the sludge char, the tar, and the producer gas during the pyrolysis in the

screw carbonizer, and then steam activation was conducted in the rotary activator. The

gravimetric tar is total amount of tar after passing carbonization and activation process.

Benzene and naphthalene among light tar are products generated during secondary pyrolysis

at the carbonizer, and some part of both tars converts to gas during the steam activation in the

activator. In addition, anthracene and pyrene were directly formed by primary pyrolysis from

the dried sludge in the carbonizer. Both tars were known as not affecting by carbonization-

activation temperature and the amount of steam feed [87].

Table 6.5 Tar concentrations from the carbonization-activation (unit: g/Nm3)

Gravimetric tar Benzene Naphthalene Anthracene Pyrene Benzonitrile Benzene-

acetonitrile

26.3 6.31 2.97 0.87 0.12 0.61 0.11

▌Producer gas characteristics

Table 6.6 shows the producer gas concentration and higher heating value from the

carbonization-activator. Major components in the producer gas were analyzed to be H2, CO,

CH4, and CO2 along with trace amount of N2 and O2. The higher heating value was 13,400

kJ/Nm3.

Table 6.6 Concentration of the producer gas and higher heating value

Producer gas (dry vol. %) Higher heating value

(kJ/Nm3) H2 CO CH4 CO2 C2H4 C2H6 O2 N2

41.2 17.3 9.5 15.4 0 0 0.5 3.3 13,400

H2 was produced by the cracking of the volatile matter generated by the pyrolysis and steam

gasification. CH4 resulted from cracking and depolymerization reactions, while CO and CO2

were produced from decarboxylation and depolymerization or the secondary oxidation of

carbon [148].

In addition, the presence of steam at high temperatures gave rise to in situ steam reforming of

the volatile matters and partial gasification of the solid carbonaceous residue, as shown in the

water gas reaction (Eq. 3.2) and tar steam gasification (Eq. 3.6).

116

Non-condensable products may also undergo gas phase reactions with each other. For

example, the CO and CH4 contents may be affected by the methane gasification and water gas

shift reactions, as shown in Eqs. 3.7 and 3.8 [149].

High temperatures were also responsible for the reduction of C2H4, C2H6 and C3H8. Some of

the typical reactions are expressed by Eqs. 3.9 and 3.10 [89].

However, it should be noted that the gas composition may not exclusively be the result of tar

cracking and the partial gasification of char due to the complicated interactions of the

intermediate products, which would probably affect the final gas composition.

2) Plasma reformer and adsorber characteristics

The plasma reformer was installed for converting generated tar from the carbonization-

activator into hydrogen-rich gas via decomposition and the steam reforming. In addition, the

fixed bed adsorber was installed for adsorption of by-passed tar from the plasma reformer.

▌Tar destruction performance

Figure 6.15 shows the results of the tar sampling at the rear section of the carbonization-

activator, the plasma reformer, and the fixed bed adsorber.

Gravimetric tar mass at the outlet of the carbonization-activator was 26.3 g/Nm3, and it was

reduced to 4.4 g/Nm3 at the plasma reformer outlet. The decomposition efficiency of the

corresponding gravimetric tar was 83.2%. For light tar, the total amount at the outlet of the

carbonization- activator 10.9 g/Nm3. The concentration was reduced to 1.3 g/Nm

3 at the outlet

of the reformer, and the decomposition efficiency of the light tar was 87.9%. Each

concentration of the light tars was found to be 0.62 g/Nm3 for benzene, 0.45 g/Nm

3 for

naphthalene, 0.14 g/Nm3 for anthracene, 0.021 g/Nm


3 for benzonitrile,

and 0.015 g/Nm3 for benzeneacetonitrile.

Decomposition of heavy tar was happened due to plasma cracking and carbon formation in

Eqs. 5.5 and 5.6 [26]. In addition, steam in producer gas from the carbonization-activator

formed excitation species as shown in Eq. 5.20, and the radicals reduced light tar and carbon-

black which produce by the reactions of the plasma cracking and carbon formation [130]. It is

remarkable that tars from the carbonization and activation should be decomposed

significantly by the plasma reformer.

Discharged residual tar from the plasma reformer was removed by the fixed bed adsorber

filled with the sludge char which produced from the carbonization-activation.

Gravimetric tar at the adsorber outlet displayed 0.5 g/Nm3, which is 88.6% of removal

efficiency. Total amount of light tar was 0.39 g/Nm3, which is corresponded to 40.5% of the

removal efficiency. The relevant concentration was 0.28 g/Nm3 for benzene, 0.09 g/Nm

3 for

naphthalene, 0.14 g/Nm3 for anthracene, 0.01 g/Nm

3 for benzonitrile, and 0.003 g/Nm

3 for

benzeneacetonitrile.

Among the residual tar, heavy tar was mostly removed at the adsorber, and particularly non-

condensed light tar that was not adsorbed like benzene was considered to be passed through

the activated carbon adsorber [63, 81]. The gravimetric tar concentration of 0.5 g/Nm3 in the

producer gas is not considered to be problematic in the operation of an IC engine, a

compressor, etc [15].

117

Figure 6.15 Gravimetric tar mass and light tar concentrations

▌Gas formation characteristics

Figure 6.16 shows the producer gas analysis sampled from the carbonization-activator, the

plasma reformer, and the fixed bed adsorber, respectively.

At the outlet of the plasma reformer, the gas concentration was found to be 50.9% for H2,

22.3% for CO, 11% for CH4, 8.7% for CO2, 0.4% for C2H2, and 0.2% for C2H4. H2, CO, and

light hydrocarbons (CH4, C2H4, and C2H6) increased and CO2 decreased, compared to the

outlet concentration of the carbonization-activator.

H2 and CO increased due to tar steam gasification of tar (Eq. 3.6) and the methane

gasification reaction (Eq. 3.7). The light hydrocarbons was converted from light tar using tar

plasma cracking reaction (Eq. 5.5) in portion and from chain reactions of Eqs. 5.18 and 5.19.

In addition, decrease in CO2 was considered to be dry reforming as shown in Eq. 6.1 [138].

CnHx + nCO2 → (x/2)H2 + 2nCO (6.1)

Figure 6.16 Producer gas concentrations at the exit of each component

Gra

vim

etr

icta

r(g

/Nm

3)

Lig

ht

tar

(g/N

m3)

0

5

10

15

20

25

30

0

1

2

3

4

5

6

7

8

Naph-thalene

Gravimetictar

Anthra-cene



Pyrolysis gasifier

Benzene

Adsorber

Plasma reformer

26.3

6.31

4.4

0.5

0.62

0.28

2.97

0.45

0.09

0.87

0.14

0.12

0.02

00

0.61

0.08

0.01

0.11

0.01

0

Ga

sco

nce

ntr

atio

n(%

)

Hig

he

rh

ea

tin

gva

lue

(kJ/N

m3)

0

10

20

30

40

50

60

0

5000

10000

15000

20000

H2

CH4

CO CO2

HHV

Pyrolsis gasifier

Adsorber

C2H

4C

2H

6O

2

Plasma reformer

N2

41.2

50.9

50.5

17.3

22.3

21.9

9.5

11

10.5

15.4

8.7

7.90

0.4

0.1

0

0.2

0

0.5

0.8

0.9

3.3

1.8

2.5

11200

13992

13482

118

According to the gas analysis results at the adsorber outlet, 50.5% of H2, 21.9% of CO,

10.5% of CH4, 7.7% of CO2, and 0.1% of C2H2 were displayed. Compared to the results at the

plasma reformer outlet, the corresponding concentration was slightly decreased, but it was not

almost adsorbed. The higher heating value was found to be 11,200 kJ/Nm3 for the producer

gas from the carbonization-activator, 13,992 kJ/Nm3 for the plasma reformer and 13,482

kJ/Nm3 for the adsorber. The increase at the plasma reformer outlet is due to increased

amount of combustible gases, particularly methane having higher heating value.

6.3.3 Process analysis for the sequential in-line carbonization-activation

system

1) Mass and energy balance for a carbonization-activator

This section describes mass and energy balance analysis of the combined carbonization-

activator at the optimal condition (Table 6.3). A combination of thermodynamics and heat

transfer approach were employed to solve the overall calculation.

The mass flow diagram is shown in Figure 6.17. The input mass flow to the combined

carbonization-activator includes the dried sludge (0.8 kg/h) and the superheated steam (0.6

kg/h) plus the flesh leak air (0.073 kg/h). The dried sludge has bone dry sludge (0.72 kg/h)

and moisture in dried sludge (0.08 kg/h). And hot combustion gas (22.1 kg/h) indirectly

heated the combined carbonization-activator. The output mass products are the sludge char

(0.283 kg/h), the producer gas (0.88 kg/h), the tar (0.11 kg/h), and the steam (0.2 kg/h).

Figure 6.17 Mass flow diagram of the combined carbonization-activator

The energy balance for the carbonization-activator was calculated, and the result are presented

in Figure 6.18.

119

▌Heat input to the combined carbonization-activator is classified into the enthalpy of the

dried sludge (chemical energy of the dried sewage sludge; latent heat of the dried sewage

sludge), the superheated steam enthalpy, and the hot combustion gas enthalpy.

I. Energy for dried sewage sludge (QCds = QCcds + QClds)

<1> Chemical energy of dried sewage sludge (QCcds)

rcds W100

sS2500)

8

0.01sO

100

sH(34000

100

sC8100QC

(6.2)

where sC , sH , sO are carbon, hydrogen, oxygen and sulfur in the dried sludge (%), and

rW is the dried sludge feed rate (kg/h).

<2> Latent heat of dried sewage sludge (QClds) [150]

r dswc

pDMwc

pllds Wt)100

W(1c

100

WcQC

(6.3)

where plc is the specific heat of water at the reference temperature (kJ/kg℃), and wcW is the

moisture in the dried sludge (%).

In Eq. 6.3, the specific heat of the dried sewage sludge (cpDM) [151] can be calculated from

Eq. 6.4.

dspDM t29.31434c (6.4)

where dst is the temperature of the dried sludge at the inlet (℃).

II . Superheated steam enthalpy (QCss)

℃)100(tc)t℃(100cLmQC spssrtplsessss

(6.5)

where ssm is the superheated steam feed rate (kg/h), eL is the latent heat at the reference

temperature (kJ/kg), plsc is the specific heat of steam at the boiling point (kJ/kg℃), pssc is the

specific heat of steam at the feed temperature (kJ/kg℃), st is the temperature of the feed

steam at the inlet (℃), and trt is the reference temperature.

III . Hot combustion gas enthalpy (QChg)

hgpihghg tcmQC (6.6)

where hgm is the actual amount of wet hot combustion gas (kg/h), pic is the specific heat of

hot combustion gas at the input temperature (kJ/kg℃), and hgt is the hot combustion gas

temperature in the combustor (kg/h).

▌Heat output from the combined carbonization-activator includes the enthalpy of the

producer gas (chemical energy of the producer gas; the sensible heat of the producer gas), the

energy of tar, the enthalpy of steam, the enthalpy of sludge char (the chemical energy of

sludge char; the sensible heat of sludge char), the heat loss in the exhaust gas and the surface

heat loss.

120

I. Energy of the producer gas (QCpg = QCcpg + QClpg)

<1> Chemical energy of the producer gas (QCcpg)

ca

dvcpcpgt273K

273KGQCQC

(6.7)

where cpQC is the higher heating value at the normal condition (kcal/Nm3), Gdv is volume

flow rate of wet producer gas (m3/h), and cat is the temperature at the outlet of the carbonization-

activator (℃).

In Eq. 6.7, the higher heating value at the normal condition ( cpQC ) can be calculated by Eq.

6.8.

100

xCH9530

100

xCO3035

100

xH3500QC 42

cp (6.8)

where 2xH , xCO , 4xCH are H2, CO, CH4 concentrations in the dry producer gas (%).

<2> Sensible heat of the producer gas (QClpg)

capgdwlpg tcGQC (6.9)

where Gdw is mass flow rate of wet producer gas (kg/h), and cpg is specific heat of the

producer gas.

II . Energy of tar (QCt) [152]

trhhvt mTQC (6.10)

where hhvT is the higher heating value of tar (kJ/kg) and trm is the gravimetric tar

production amount (kg/h).

III . Enthalpy of steam (QCs)

)t(tmcQC rtcasspss (6.11)

where psc is the specific heat of steam at the outlet temperature (kJ/kg℃), and ssm is the

amount of steam in the produce gas (kg/h).

IV. Energy of sludge char (QCsc = QCcsc + QClsc)

<1> Chemical energy of sludge char (QCcsc)

sccsc m100

cS2500)

8

0.01cO

100

cH(34000

100

cC8100QC

(6.12)

where cC , cH , cO , cS are carbon, hydrogen, oxygen and sulfur concentrations in the

sludge char (%).

<2> Sensible heat of sludge char (QClsc) [153]

)t(tcmQC rtscscsclsc (6.13)

where scm is the mass flow rate of sludge char at the outlet (kg/h), scc is the specific heat of

sludge char (kJ/kg℃), and sct is the temperature of sludge char (℃).

V. Heat loss in exhaust gas (QChgl)

121

hgepocwhgl tcGQC (6.14)

where cwG is the actual amount of wet combustion gas (kg/h), and poc is the specific heat of

the exhaust gas at the exit (kJ/kg℃).

VI. Surface heat loss (QCsl)

The surface heat loss like the radiation loss can be calculated by the difference between the

total input energy (Eqs. 6.2~6.6) and the total heat output (Eqs. 6.7~6.14).

hglscstrpghgssdssl QCQCQCQCQCQCQCQCQC

(6.15)

Figure 6.18 Diagram of the energy balance for the carbonization-activator

To show the performance of the combined carbonization-activator, the net cold gas efficiency,

the hot gas efficiency and the thermal efficiency are defined as Eqs. 6.16, 6.17 and 6.18,

respectively. The net cold gas efficiency, the hot gas efficiency and the thermal efficiency

were 35.2%, 50.7% and 51.2%, respectively.

100inputheatTotal

gasproducer theofenergy ChemicalefficiencygascoldNet

100QCQCQC

QC

hgssds

cpg

(6.16)

100inputheatTotal

tar theof Energy gasproducer theofEnergy efficiencygasHot

100QCQCQC

QCQC

hgssds

trpg

(6.17)

100inputheatTotal

steam theofEnthalpy tar theof Energy gasproducer theofEnergy efficiencyThermal

100QCQCQC

QCQCQC

hgssds

strpg

(6.18)

122

Table 6.7 shows the results for the calculation of the heat balance through Eqs. 6.2 to 6.18.

Table 6.7 Calculated energy balance of the combined carbonization-activator


heat input Symbol

Heat input to a combined

carbonization-activator kJ/h 50,616.74 100 hgssds QCQCQC

I. Energy for dried sludge

<1> Chemical energy of dried

sewage sludge

<2> Latent heat of dried

sewage sludge

kJ/h

kJ/h

19,029.55

28.16

37.6

0.1

cdsQC

ldsQC

Sub total I kJ/h 19,057.71 37.7 dsQC

II. Superheated steam enthalpy kJ/h 2,078.03 4.1 ssQC

III. Hot combustion gas

enthalpy kJ/h 2,9481 58.2 hgQC

Total output energy

(= I + II + III + IV + V + VI) kJ/h 50,616.74 100

slhglsc

strpg

QCQCQC

QCQCQC

Heat output for hot gas

(= I + II) kJ/h 25710.92 50.7

Cpg+ C

tr;

Hot gas efficiency


(= I + II + III) kJ/h 24,879.82 51.2 strpg QCQCQC ;

Thermal efficiency

I. Energy of producer gas

<1> Chemical energy of

producer gas

<2> Sensible heat of producer

gas

kJ/h

kJ/h

17,796.92

3,514

35.2

6.9

cpgQC ; Net cold gas efficiency

lpgQC

Sub total I kJ/h 21,310.92 42.1 pgQC

II. Energy of tar kJ/h 4400 8.7 trQC

III. Enthalpy of steam kJ/h 181.72 0.4 sQC

IV. Energy of sludge char

<1> Chemical energy of sludge

char

<2> Sensible heat of sludge

char

kJ/h

kJ/h

8,681.76

52.98

17.2

0.1

cssQC

lscQC

Sub total IV kJ/h 8,734.74 17.3 scQC

V. Heat loss in exhaust gas kJ/h 5,977.61 11.8 hglQC

VI. Surface heat loss kJ/h 10,011.75 19.7 )QCQCQCQC

QCQCQCQC( QC

hglscstr

pghgssdssl

Figure 6.19 represents the energy flow diagram for the carbonization-activator.

123

The heat input energy was supplied to the combined carbonization-activator by the hot

combustion gas (58.2%) indirectly. And the superheated steam (4.1%) was directly injected

into the rotary activator for the steam activation. The dried sewage sludge (37.7%), which is

produced by the rotary drum dryer, was fed to the screw carbonizer, having the chemical

energy (37.6%) and the latent heat (0.1%).

The heat output energy includes the producer gas (42.1%) having the chemical energy (35.2%)

and the sensible heat (6.9%), the sludge char (17.3%) having the chemical energy (17.2%)

and the sensible heat (0.1%), the tar (8.7%) having chemical energy, the steam enthalpy

(0.4%), and the heat loss from the combined carbonization-activator is the surface heat loss

(19.7%) and the exhaust gas loss (11.8%) for the combustion hot gas.

Figure 6.19 Diagram of the energy flow for the carbonization-activator

In summary, the purpose of this study is to convert the waste sewage sludge to energy (i.e.,

clean producer gas fuel) and resources (i.e., high quality sludge char).

In the sight of energy use for the waste sewage sludge, the combined carbonization-activator

produces useful energy of 35.2% (i.e., the net cold gas efficiency is 35.2%) as the producer

gas fuel which will be used as the fuel for gas engines, gas turbines, fuel cells, etc.). The

conversion of the tar into light gas gives a possibility as the usage in fuel energy. As for this

case, the combined carbonization-activator produces the energy of 50.7% (i.e., the hot gas

efficiency is 50.7%). And the thermal efficiency considered the steam enthalpy is 51.2%.

2) Mass and energy balance for a plasma reformer

This section describes mass and energy balance analysis of the gliding arc plasma reformer

(GAPR) at the optimal condition like the combined carbonization-activator.

The mass flow diagram is shown in Figure 6.20. The input mass flow to the plasma reformer

includes the producer gas (0.88 kg/h), the tar (0.11 kg/h), and the steam (0.2 kg/h); The output

mass products are the reforming gas (1.008 kg/h), the tar (0.017 kg/h), and the steam (0.165

kg/h).

124

Figure 6.20 Mass flow diagram of the gliding arc plasma reformer

The energy balance for the plasma reformer was calculated, and the result is presented in

Figure 6.21.

▌Heat input to the plasma reformer is classified into the energy of the producer gas, the

energy of the tar, and the enthalpy of the steam. The energy of the producer gas (21,310.92

kJ/h) includes the chemical energy of the producer gas (17796.92 kJ/h) and the sensible heat

of the producer gas (3514 kJ/h). The energy of the tar and the enthalpy of the steam are 4,400

kJ/h and 181.72 kJ/h, respectively. Each value is the same that at the outlet of the

carbonization-activator. In addition, the input electric energy was 1008 kJ/h (i.e., 0.28 kW) to

generate plasma discharge between three gliding electrodes.

▌Heat output from the plasma reformer includes the energy of the reforming gas (chemical

energy and sensible heat), the energy of the tar, the enthalpy of the steam, and the surface heat

loss.

I. Energy of the reforming gas (QPrg = QPcrg + QPlrg)

<1> Chemical energy of the reforming gas (QPcrg)

pr

dvcrcrgt273K

273KGQPQP

(6.19)

where crQP is the higher heating value at the normal condition (kcal/Nm3), dvG is the

volume flow rate of wet reforming gas (m3/h), and prt

is the temperature at the outlet of the

plasma reformer (℃).

In Eq. 6.19, the higher heating value at the normal condition (QPcrg) can be calculated by Eq.

6.20.

100

yCH9530

100

yCO3035

100

yH(3050QP 42

cr

)100

HyC16820

100

HyC15280 6242 (6.20)

where 2yH , yCO , 4yCH , 42HyC and 62HyC are the concentrations of H2, CO, CH4,

C2H4, and C2H6 in the dry reforming gas (%).

<2> Sensible heat of the reforming gas (QPlrg)

prrgwglrg tcmQP (6.21)

125

where wgm is the mass flow rate of wet reforming gas (kg/h), and crg is the specific heat of

the reforming gas.

The mass flow rate of the wet reforming gas can be calculated by Eq. (6.22).

mwg= mrg+ mrs (6.22)

where mrg is the mass flow rate of the dry reforming gas (kg/h), and mrs is the mass flow

rate of the steam (kg/h).

II . Energy of tar (QPtr)

Ptr=Thhv mptr (6.23)

where Thhv is the higher heating value of the tar (kJ/kg), and mptr is the mass flow rate of the

tar (kg/h).

III . Enthalpy of steam (QPrs)

Prs=cps mrs (tpr - trt) (6.24)

where cps is the specific heat of steam at the outlet temperature of the plasma reformer

(kJ/kg℃), tpr is the temperature at the outlet of the plasma reformer (℃).

IV. Surface heat loss (QPsl)

The surface heat loss like the radiation surface loss can be calculated by the difference

between the total input energy (which is the same value at the outlet of the carbonization-

activator) and the total heat output (calculated from Eqs. 6.19~6.24).

Psl= Cpg + Cct + Cps + Ppe- Prg - Ppt - Prs (6.25)

Figure 6.21 Diagram of energy balance for the gliding arc plasma reformer

To show the performance of the gliding arc plasma reformer, the cold gas efficiency, the net

cold gas efficiency, the hot gas efficiency, and the thermal efficiency are defined as Eqs.

6.26~6.29, respectively. The cold gas efficiency is 86%; the net cold gas efficiency is 82.2%;

the hot gas efficiency is 99.1%; the thermal efficiency is 99.6%.

Cold gas efficiency = Chemical energy of the reforming gas

Chemical energy of feedstock 100

126

100QCQC

QP

trpg

crg

(6.26)

Net cold gas efficiency = Chemical energy of the reforming gas

Total energy input 100

100QP+QC+QCQC

QP

pepstrpg

crg

(6.27)

Hot gas efficiency = Energy of the reforming gas+Energy of the tar


100QP+QC+QCQC

QPQP

pepstrpg

trrg

(6.28)

Thermal efficiency=Energy of the reforming gas+Energy of the tar+Enthalpy of the steam


100QP+QC+QCQC

QPQPQP

pepstrpg

rstrrg

(6.29)

Table 6.8 shows the results of the calculation of the heat balance through Eqs. 6.19 to 6.29.

Table 6.8 Calculated energy balance for the gliding arc plasma reformer


heat input Symbol

Heat input to the plasma reformer kJ/h 26,900.64 100

I. Energy of producer gas

(i) Chemical energy of producer gas

(ii) Sensible heat of producer gas

kJ/h

kJ/h

17,796.92

3,514

66.2

13.1

Ccpg

Clpg

Sub total I kJ/h 21,310.92 79.3 Cpg

II. Energy of tar kJ/h 4,400 16.3 Ctr

III. Enthalpy of steam kJ/h 181.72 0.7 Cs

IV. Input electric energy kJ/h 1,008 3.7 Ppe

Total output energy

(= I + II + III + IV ) kJ/h 26,900.64 100

Energy output from the plasma

reformer (= I + II + III ) kJ/h 26,788.44 99.6

Prg

Ptr

Prs

;

Thermal efficiency


(= I + II) kJ/h 26,660.1 99.1

Prg+ P

tr;

Hot gas efficiency

I. Energy of reforming gas

(i) Chemical energy of reforming gas

(ii) Sensible heat of reforming gas

kJ/h

kJ/h

22,106.39

3,873.71

82.2

14.4

Pcrg

; Net cold gas efficiency

Plrg

Sub total I kJ/h 25,980.1 96.6 Prg

II. Energy of tar kJ/h 680 2.5 Ptr

III. Enthalpy of steam kJ/h 128.34 0.5 Prs

VI. Surface heat loss kJ/h 112.2 0.4 P

sl (= Cpg + Ctr + Cps

+ Ppe- Prg - Ptr - Prs

127

Figure 6.22 represents the energy flow diagram for the gliding arc plasma reformer.

The heat input energy was supplied to the plasma reformer by the producer gas of 79.3%

(chemical energy of 66.2%; sensible heat gas of 13.1%), the tar of 16.3%, and the steam of

0.7% from the combined carbonization-activator. In addition, the input electric energy of 3.7%

was supplied from the power supply.

The heat output energy includes the reforming gas (96.6%) having the chemical energy

(82.2%) and the sensible heat (14.4%), the tar (2.5%), the steam (0.5%), and the surface heat

loss (0.4%).

Figure 6.22 Diagram of the energy flow for the gliding arc plasma reformer

The target of the plasma reformer is to destruct the tar, particularly heavy tar, increasing the

heating value of the reforming gas by the conversion of this heavy tar into light gases.

In the plasma reformer, the energy of the producer gas (79.3%) and the tar (16.3%) were

converted to the chemical energy of the reforming gas (82.2%), showing that the cold gas

efficiency was 86%. Particularly, the converted energy of input tar into the reforming light

gas is 13.8%. It should be useful to increase the heating value of the reforming gas.

Regarding to the total input energy, the plasma reformer can produce useful energy of 82.2%

(i.e., the net cold gas efficiency is 82.2%). The conversion of the tar into light gas makes this

a useful fuel energy. As for this case, the plasma reformer can produce the energy of 99.1%

(i.e., the hot gas efficiency is 99.1%). In addition, the use of the steam energy gives higher

thermal efficiency of 99.6% due to low surface heat loss (0.4%) by the well wall insulation.

Through the energy analysis including the plasma reformer, the input electric energy should

be converted to the primary energy for the comparison with other systems such as using

thermal decomposition like high temperature steam-catalytic reformer, etc.

3) Mass and energy balance for an adsorber

This section describes the mass and energy balance analysis of the fixed bed adsorber at the

optimal condition.

The mass flow diagram is shown in Figure 6.23. The input mass flow to the adsorber includes

the reforming gas (1.008 kg/h), the tar (0.017 kg/h), and the steam (0.165 kg/h); The output

128

mass products are the cleaned gas (1.008 kg/h), the bypass tar (0.001 kg/h), and the

condensed water (0.165 kg/h). In addition, the adsorbed tar in the adsorber was 0.016 kg/h.

Figure 6.23 Mass flow diagram of the fixed bed adsorber

The energy balance for the fixed bed adsorber was calculated, and the result is presented in

Figure 6.24.

▌Heat input to the fixed bed adsorber is classified into the energy of the reforming gas, the

energy of the tar, and the enthalpy of the steam. The energy of the reforming gas includes the

chemical energy (22106.39 kJ/h) and the sensible heat (3873.71 kJ/h). The energy of the tar

and the enthalpy of the steam are 680 kJ/h and 128.34 kJ/h, respectively.

▌Heat output from the adsorber includes the energy of the cleaned gas (the chemical energy

and sensible heat), the energy of the tar, the enthalpy of the steam, condensed water heat

loss and the cooling heat loss.

I. Energy of the cleaned gas (QAcg = QAccg + QAlcg)

<1> Chemical energy of cleaned gas (QAccg)

ad

avccccgt273K

273KGQAQA

(6.30)

where ccQA is the higher heating value at the normal condition (kcal/Nm3), avG is the

volume flow rate of the wet cleaned gas (m3/h), and adt is the temperature at the outlet of the

adsorber (℃).

In Eq. 6.30, the higher heating value at the normal condition ( ccQA ) can be calculated by Eq.

6.31.

)100

HzC15280

100

zCH9530

100

zCO3035

100

zH(3050QA 4242

cc (6.31)

where zH2 , zCO , zCH4, and zC2H4 are the concentrations of H2, CO, CH4, C2H4 in the dry

cleaned gas (%).

129

<2> Sensible heat of cleaned gas (QAlcg)

Alcg= mcg ccg tad (6.32)

where m g is the mass flow rate of the wet cleaned gas (kg/h) and ccg is the specific heat of

the cleaned gas.

II . Energy of tar (QAtr)

Atr=Thhv matr (6.33)

where Thhv is the higher heating value of the tar (kJ/kg) and m tr is the mass flow rate of the

tar (kg/h).

III . Adsorbed tar loss (QAtl)

Atl=Thhv mdtr (6.34)

where mdtr is the mass flow rate of the adsorbed tar (kg/h)

IV. Condensed water heat loss (QAcl)

Acl=mrs crs (tpr-twt) (6.35)

where mrs is the steam flow rate from the reforming gas (kg/h), crs is the specific heat of the

steam at the outlet of the plasma reformer (kJ/kg℃), tpr is the temperature at the outlet of the

plasma reformer (℃), and twt is the temperature of the water trap (℃)

V. Cooling heat loss (QAhl)

The cooling heat loss is calculated by the difference between the total input energy (which is

the same value at the outlet of the plasma reformer) and the total heat output (calculated from

Eqs. 6.30~6.35).

Ahl= Prg + Ptr + Prs - Acg - Atr - Atl - Acl (6.36)

Figure 6.24 Diagram of the energy balance for the fixed bed adsorber

To show the performance of the fixed bed adsorber, the net cold gas efficiency, the hot gas

efficiency, and the thermal efficiency are defined as Eqs. 6.37~6.39, respectively. The net

cold gas efficiency is 78.8%; the hot gas efficiency is 80%; the thermal efficiency is 97.1%.

130

Net cold gas efficiency = Chemical energy of the cleaned gas

Total heat input 100

100QP+QPQP

QA

rstrrg

ccg

(6.37)

Hot gas efficiency = Energy of the cleaned gas+Energy of the tar


100QP+QPQP

QAQA

rstrrg

trcg

(6.38)

Thermal efficiency = Energy of the cleaned gas+Energy of the tar+Cooling heat loss


100QP+QPQP

QAQAQA

rstrrg

hltrcg

(6.39)

Table 6.9 shows the results of the calculation of the heat balance through Eqs. 6.30 to 6.39.

Table 6.9 Calculated energy balance for the fixed bed adsorber


heat input Symbol

Heat input to the adsorber kJ/h 26788.44 100 r tr r

I. Energy of reforming gas

(i) Chemical energy of reforming

gas

(ii) Sensible heat of reforming

gas

kJ/h

kJ/h

22106.39

3873.71

82.5

14.5

Pcrg

Plrg

Sub total I kJ/h 25980.1 97.0 Prg

II. Energy of tar kJ/h 680 2.5 Ptr

III. Enthalpy of steam kJ/h 128.34 0.5 Prs

Total output energy

(= I + II + III + IV + V) kJ/h 26788.44 100

Heat output from an adsorber

(= I + II + V) kJ/h 26006.48 97.1

g tr hl;

Thermal efficiency


(= I + II) kJ/h 21431.04 80

g tr;

Hot gas efficiency

I. Energy of cleaned gas

(i) Chemical energy of cleaned gas

(ii) Sensible heat of cleaned gas

kJ/h

kJ/h

21116.32

274.72

78.8

1.1

Ac g; Net cold gas efficiency

l g

Sub total I kJ/h 21391.04 79.9 A g

II. Energy of tar kJ/h 40 0.1 Atr

III. Adsorbed tar loss kJ/h 640 2.4 Atl

IV. Condensed water heat loss kJ/h 141.96 0.5 A l

V. Cooling heat loss kJ/h 4575.44 17.1 Ahl (= Prg + Ptr + Prs - Acg - Atr - Atl - Acl

131

Figure 6.25 represents the energy flow diagram for the fixed bed adsorber.

The heat input energy was supplied to the adsorber by the reforming gas of 97% (the chemical

energy of 82.5%; the sensible heat of 14.5%), the tar of 2.5%, and the steam of 0.5% from the

plasma reformer.

The heat output energy includes the cleaned gas (79.9%) having the chemical energy (78.8%)

and the sensible heat (1.1%), and the bypass tar (0.1%). The heat loss in the water trap

includes the condensed water heat loss (0.5%) and the cooling heat loss (17.1%). In addition,

the adsorbed tar loss was 2.4% due to tar adsorption.

Figure 6.25 Diagram of the energy flow for the fixed bed adsorber

The target of the adsorber which uses the sludge char as an adsorbent, is to remove the

residual tar from the plasma reformer.

Regarding to the total input energy, the adsorber can purify the reforming gas as a useful

energy of 78.8% (i.e., the net cold gas efficiency is 78.8%) and the clean producer gas will be

used for end-use devices. If the bypass tar would be light aromatic tar (like benzene, toluene,

etc.), it can be also used as the fuel energy. In this case, the adsorber can produce the clean

producer gas with the energy of 80% (i.e., the hot gas efficiency is 80%). In addition, the use

of the cooling heat at the water trap can give higher energy usage for the system, showing the

thermal efficiency of 97.1%.

4) Performance analysis in view of the total energy balance for a sequential

in-line treatment system

The energy balance for the sequential carbonization-activation system was calculated, and the

result is presented in Table 6.10.

The heat input to the combined carbonization-activator is classified into the enthalpy of the

dried sewage sludge (19057.71 kJ/h) including the chemical energy of the dried sewage

sludge (19029.55 kJ/h) and the latent heat of the dried sewage sludge (28.16 kJ/h), the

superheated steam enthalpy (2078.03 kJ/h), and the hot combustion gas enthalpy (29481 kJ/h).

In addition, the input electric power (1008 kJ/h) was supplied to the gliding arc plasma

reformer.

The heat output from the adsorber includes the energy of the cleaned producer gas (21391.04

kJ/h) including the chemical energy of the cleaned producer gas (21116.32 kJ/h) and the

sensible heat of the cleaned producer gas (274.72 kJ/h), the energy of the tar (40 kJ/h), energy

of the sludge char (8734.74 kJ/h) including the chemical energy (8681.76 kJ/h) and the

sensible heat (52.98 kJ/h). The total energy loss through the system (including the combined

132

carbonization-activator surface heat loss, the plasma reformer surface heat loss, the cooling

heat loss, the adsorbed tar loss, and the condensed water heat loss) was 15481.35 kJ/h.

Table 6.10 Calculated total energy balance for the sequential carbonization-activation system


heat input Symbol

Heat input to carbonization-

activation system kJ/h 51624.74 100

I. Energy for dried sludge

(i) Chemical energy of dried

sewage sludge

(ii) Latent heat of dried sewage

sludge

kJ/h

kJ/h

19029.55

28.16

36.9

0.1

Ccds

Clds

Sub total I kJ/h 19057.71 37.0 Cds

II. Superheated steam enthalpy kJ/h 2078.03 4.0 Css

III. Hot combustion gas enthalpy kJ/h 29481 57.1 Chg

IV. Energy of electric power kJ/h 1008 1.9 Cpe

Heat output from adsorber kJ/h 51624.74 100


(= I + II) kJ/h 21431.04 41.5

Acg+ A

tr;

Hot gas efficiency

I. Energy of cleaned producer gas

(i) Chemical energy of cleaned

producer gas

(ii) Sensible heat of cleaned

producer gas

kJ/h

kJ/h

21116.32

274.72

40.9

0.5

Accg

;

Net cold gas efficiency

Alcg

Sub total I kJ/h 21391.04 41.4 Acg

II. Chemical energy of tar kJ/h 40 0.1 Atr

III. Energy of sludge char

(i) Chemical energy of sludge char

(ii) Sensible heat of sludge char

kJ/h

kJ/h

8681.76

52.98

16.8

0.1

Ccsc

Clsc

Sub total III kJ/h 8734.74 16.9 Csc

IV. Heat loss in exhaust gas kJ/h 5977.61 11.6 Chgl

V.Heat loss in each component

(i) Combined carbonization-

activator surface heat loss

(ii) Plasma reformer surface

heat loss

(iii) Cooling heat loss

(iii) Adsorbed tar loss

(iv) Condensed water heat loss

kJ/h

kJ/h

kJ/h

kJ/h

kJ/h

10011.75

112.2

4575.44

640

141.96

19.4

0.2

8.9

1.2

0.3

Csl

Psl

Ahl

Atl

Acl

Sub total V kJ/h 15481.35 30.0

Tsl

Tsl

Csl

Psl

Ahl

Atl A

cl

133

The definitions of the efficiencies shown in Table 6.10 are expressed as Eqs. 6.40 and 6.41,

respectively. The net cold gas efficiency and the hot gas efficiency are 40.9% and 41.5%

respectively.

Net cold gas efficiency = Chemical energy of the cleaned producer gas


100QP+QC+QCQC

QA

pehgssds

ccg

(6.40)

Hot gas efficiency = Energy of the cleaned producer gas+Energy of the tar


100QP+QC+QCQC

QAQA

pehgssds

trcg

(6.41)

Figure 6.26 represents the energy flow diagram for the sequential carbonization-activation

system.

The heat input energy was supplied to the combined carbonization-activator by the dried

sewage sludge of 37.0% (the chemical energy of 36.9%; the latent heat of 0.1%), the

superheated steam of 4.0%, and the hot combustion gas of 57.1%. And the input electric

energy to the plasma reformer was 1.9%.

The heat output energy includes the cleaned producer gas (41.4%) having the chemical energy

(40.9%) and the sensible heat (0.5%), and the bypass tar (0.1%). The heat loss in the water

trap includes the condensed water heat loss (0.3%) and the cooling heat loss (8.9%). The

sludge char from the adsorber has the heat loss including the sludge char (16.8%) and the

adsorbed tar loss (1.2%) due to tar adsorption. In addition, the heat loss from the combined

carbonization-activator was the sensible heat of the sludge char (0.1%) and heat loss in the

exhaust gas (11.6%). And the total surface heat loss from each component was 19.6%.

The main aim of this system is to treat the waste sewage sludge from the waste water

treatment plant eco-friendly. In addition, the production for high quality producer gas and

sludge char should be achieved. The sludge char is used for the gas purification in the

adsorber.

Figure 6.26 Diagram of the energy flow for the sequential carbonization-activation system

134

Regarding to the total input energy, the sequential carbonization-activation system produces

clean gas fuel with the energy of 40.9% (i.e., the net cold gas efficiency is 40.9%) which will

be applied for end-use devices. If the bypass tar should be light aromatic tar (like benzene,

toluene, etc.), it can be also used as the fuel energy. As for this case, the system will produce

the energy of 41.5% (i.e., the hot gas efficiency is 41.5%).

The slight different value between the net cold gas efficiency and the by gas efficiency means

that this system well achieves clean gas production by tar conversion and removal by the after

treatment technology (i.e., plasma reformer and adsorber).

But the heat loss like the surface heat loss, the cooling heat loss, the exhaust heat loss share

large portion of the output energy. In addition, the sludge char including adsorbed tar was not

used for the thermal exergy in this system. Therefore, the improvement of the system

efficiency should need the usage of the lost heat as well as the effective utilization of the tar-

adsorbed sludge char as a fuel.

For the improvement of the system efficiency, two scenarios are suggested as shown in

Figures 6.27 and 6.28.

The energy flow diagram in the Figure 6.27 represents the usage as a fuel for the sludge char

including the adsorbed tar from the fixed bed adsorber. The net cold gas efficiency and the

hot gas efficiency defined by Eqs. 6.42 and 6.43 are 49.9% and 50.7%, respectively. The

adsorbed sludge char (18%; 9,322 kJ/h) will be used as a solid fuel instead of the LPG fuel

which generates the hot combustion gas for the combined carbonization-activator. That is, the

energy content of the tar-adsorbed sludge char will be recovered for reduction of the LPG fuel

usage.

100inputheat Total

gasproducer cleaned theofenergy ChemicalefficiencygascoldNet (6.42)

100inputheat Total

tar theofEnergy gasproducer cleaned theofEnergy efficiency gasHot

(6.43)

Figure 6.27 Flow diagram including the usage of the tar-adsorbed sludge char

135

Figure 6.28 represents the improved system including the adsorbed sludge char usage and the

lost heat including the exhaust heat loss and the cooling heat loss. The net cold gas efficiency

and the hot gas efficiency defined by Eqs. 6.42 and 6.43 are 66.5% and 67.5%, respectively.

The heat loss of the exhaust gas (11.6%; 5978 kJ/h) will be recovered in the combined

carbonization-activator, and the cooling heat from the water trap will be used for increasing

the enthalpy for the plasma reformer. This energy recuperation should maximize the system

efficiency as can be seen from the improvement of the net cold gas efficiency and the hot gas

efficiency.

Compared to the original sequential carbonization-activation system, the net cold gas

efficiency and the hot gas efficiency should be improved to 25.6% and 26% for this

recuperative system. In addition, further improvement for the system efficiency can be

expected by changing the thermal insulation to reduce the surface heat loss in each component,

particularly the combined carbonization-activator.

Figure 6.28 Flow diagram including the usage of the tar-adsorbed sludge char and the lost

heat

6.4 Summary

A combined carbonization-activator was developed for conversion of waste sewage sludge to

energy and resources, and the parametric study was conducted to show the best operating

characteristics. And mass and energy balances were conducted to verify the thermal

performance; A sequential in-line system for production of high quality sludge char and

producer gas was suggested and verified its performance.

▌Combined carbonization-activator; The combined carbonization-activator, combined with

pyrolysis and the steam gasification processes, was developed for the production of an

activated sludge char and producer gas.

To know the production characteristics of the carbonization-activator, parametric researches

were conducted on the steam feed rate, the activator temperature and the sludge moisture

content. The results were as follows; First, micro-pores were well-developed with the steam

136

feed, having best condition of 10 mL/min. However, the higher heating value of the producer

gas was decreased. The performance of benzene adsorption showed a maximum saturation

point after 45 minutes, with 140 mg/g adsorbed. Second, the pore development in the

activated sludge char improved with increasing the activator temperature. The higher heating

value of the producer gas also increased. Third, the heating value of the gas decreased with

increasing the sludge moisture content. In addition, the moisture content of about 11%

displayed the largest micro-pore development in the activated sludge char. Through the

parametric research, the optimal operating conditions were taken from the viewpoints of the

highest adsorption by the activated char and the heating value of the producer gas.

▌Sequential in-line carbonization-activation system; For energy and resource utilization of

dried sewage sludge, an integrated system with a sequential in-line connection of the

carbonization-activator, the plasma reformer, and the fixed bed adsorber was developed. The

plasma reformer was set to improve the producer gas yield by destructing tar released from

the carbonization-activator. The fixed bed adsorber filled with the sludge char produced from

the carbonization-activator was installed for adsorption of un-treated tar. The carbonization-

activator produced sludge char, tar and gas.

Through the system analysis, the results for each component are shown as bellows; For the

carbonization-activator, the sludge char showed 98.1 m2/g of the specific surface area and 6.4

nm of the mean pore size, which had a good distribution of micropore and mesopore with a

superior adsorption rate for light PAH tar. The concentrations of the gravimetric tar and the

total light tar were 26.3 g/Nm3 and 10.9 g/Nm

3, respectively. The analyzed light tar was in the

order of benzene, naphthalene, benzonitrile, benzeneacetonitrile, anthracene and pyrene. The

produced gas was composed of hydrogen, carbon monoxide, methane, and carbon dioxide.

The plasma reformer displayed 83.2% of the decomposition efficiency with 4.4 g/Nm3 of

gravimetric tar at the outlet due to the tar cracking and the steam reforming reaction. And the

total amount of light tar was 1.3 g/Nm3. Among the reforming gas, concentration of hydrogen,

carbon monoxide, and methane was increased.

Gravimetric tar at the outlet of the adsorber was 0.5 g/Nm3, which was 88.6% of the removal

efficiency. And the total amount of light tar was 0.39 g/Nm3. Gas analysis results at the exit

showed 50.5% H2, 21.9% CO, 10.5% CH4, 7.9% CO2, and 0.1% C2H4 with the higher heating

value of 13,482 kJ/Nm3.

▌Process analysis for the sequential in-line system; The calculations of the mass and

energy balance for each component (the combined carbonization-activator; the gliding arc

plasma reformer; the fixed bed adsorber) were conducted. As for the combined carbonization-

activator, the net cold gas efficiency and the hot gas efficiency were 35.2% and 50.7%,

respectively. For the plasma reformer, the cold gas efficiency was 86.0%; the net cold gas

efficiency was 82.2%; the hot gas efficiency was 99.1%; the thermal efficiency was 99.6%. In

case of the adsorber, the net cold gas efficiency was 78.8%; the hot gas efficiency was 80%;

the thermal efficiency was 97.1%.

The sequential carbonization-activation system produced clean gas fuel with the energy of

40.9% (i.e., the net cold gas efficiency was 40.9%). As the bypass tar should be used as the

fuel energy, the system produced the energy of 41.5% (i.e., the hot gas efficiency was 41.5%).

Furthermore, for the improvement of the system efficiency, two scenarios were suggested,

and the verification of the system efficiency improvement was performed, showing the best

process with the cold gas efficiency of 66.5% and the hot gas efficiency of 67.5%.

137

Chapter 7

Conclusion This thesis is to propose a sequential treatment system for waste sewage sludge, and to verify

the performance of the system for production of high quality producer gas and sludge char.

The sludge treatment system integrates as an in-lined process composed of the rotary drum

dryer, the combined carbonization-activator, the plasma reformer and the fixed bed tar

adsorber. In addition, the plasma-catalyst reformer was designed for producing hydrogen-rich

gas from digesters’ biogas.

▌ In Chapter 2, a novel rotary drum dryer was developed for drying dewatered sludge

coming from a centrifuge.

The developed dryer was a new design, particularly in regards to the rotary kiln body (the

deflector, the pickup flights, the internal screw vane, the cylinder core) and the inside rotating

body (the knife-like blades, the fork-like stirrers, the fan-like blades). The newly designed

parts can improve the drying efficiency and the energy efficiency, with lower volatile

compounds production compared with conventional rotary dryers.

For verifying the effectiveness of sludge drying, parametric screening studies were conducted

by varying the rotating drum temperature, the sludge residence time, and the dryer load; The

best operating conditions were found to be 255℃ of the rotating drum temperature, 17

minutes of the sludge residence time, and 55 kg/m3·h of the dryer load. The average diameter

of the dried sludge created was about 8 mm and the weight reduction was 80%. The drying

efficiency and the moisture content in the dried sludge were 84.8 and 12.4%, respectively.

And the thermal efficiency was 73.8%, and the specific energy consumption was 3.49 MJ/kg

of water which is mostly the lowest value compared with other typical dryers.

The dried sludge produced from the novel rotary drum dryer was used as a feedstock of the

combined carbonization-activator.

▌ In Chapter 3, a batch-type wire-mesh reactor was used to find the best method for

producing high quality sludge char and producer gas simultaneously. Comparative analysis on

the formation characteristics of products such as gas, tar, and char were evaluated for each

case (i.e., the pyrolysis, the steam gasification, and the carbonization-activation).

The pyrolysis without the steam feed formed 43.9% of sludge char, 22.3% of tar, and 33.8%

of producer gas. The total amount of the producer gas was 11.5 L. The steam gasification was

achieved by continuously feeding steam from the beginning of the process. Product was

39.2% of sludge char, 23.5% of tar, and 37.3% of producer gas. The total amount of producer

gas was 20.1 L. The carbonization-activation was achieved by the pyrolysis up to 500℃ and

then gasification by feeding steam. The product was 40.1% of sludge char, 22.7% of tar, and

37.2% of gas. The total amount of the producer gas was 16.5 L.

The development of mesopore in the sludge char well adsorbed condensable tar, while the

non-condensible tar passed the sludge char bed. Therefore, the sludge char could be effective

138

for condensable tar reduction in the producer gas produced in the pyrolysis and/or gasification.

In summary, it was found to be effective to form organic volatilization by the primary

pyrolysis and to producer higher porosity sludge char and clean producer gas by the steam

activation in the secondary gasification. That is, the carbonization-activation was found to be

the best option among the three cases.

▌ In Chapter 4, the gliding arc plasma reformer (GAPR) was designed and verified its

performance for biogas to show the catalytic effect in hydrogen rich gas production. The

GAPR was combined with the catalyst reactor.

The parametric screening studies were carried out by changing the steam feed rate (i.e., the

steam/carbon ratio), the catalyst bed temperature, the total gas feed rate, the input electric

power, and the biogas content for the variables that affect reforming of the biogas in the

GAPR.

And the optimal operating conditions were shown for hydrogen rich gas production. The

optimal operating conditions and their results showed the concentrations of 62% of H2, 8% of

CO, 27% of CO2, and 0% of CH4 on the basis of the steam/carbon ratio of 3, the catalyst bed

temperature of 700℃, the total gas feed rate of 16 L/min, the input electric power of 2.4 kW,

and the biogas content of 6:4 (CH4:CO2). Also, the CH4 conversion rate was almost 100%,

and the H2 yield and the H2 selectivity were 59% and 31% respectively. At these conditions,

the energy efficiency was 53%, and the specific energy requirement was 289 kJ/mol.


conversion rate, and maintained optimal operating status regardless of the for gas property. In

addition, it is open to the application of various kinds of light gas reforming and tar

destruction for pyrolysis and/or gasification gases.

▌ Chapter 5, the GAPR developed in Chapter 4 was used for tar decomposition in the

following 3 cases.

First, benzene was used as light aromatic tar and anthracene was used as a representative light

PAH tar. Experiments were performed on the parameters that affect the tar decomposition

efficiency, and the optimal operation conditions were presented. The operating parameters

investigated were the steam flow rate, the input benzene concentration, the total gas feed rate,

and the specific energy input. Also the effects of design factors such as the nozzle diameter,

the electrode gas, the electrode length, and the electrode shape were investigated. For the

optimal design, the ratio of the electrode gap to the nozzle diameter must be 1 or higher. In

addition, an electrode type of the Arc 1showed the highest decomposition and energy

efficiencies because it ensured a sufficient plasma discharge column.

Second, to verify the performance of the plasma reformer for real tar, the continuous-type

screw pyrolyzer was designed and used for tar removal test at the optimal conditions. Tar was

sampled and analyzed for the gravimetric tar and wet group light tars. The gravimetric tar

mass was significantly reduced to 3.74 g/Nm3 at the outlet of the reformer from 18.02 g/Nm

3

at the inlet of the reformer. The removal efficiency was 79.2%, accordingly.

139

Lastly, an externally oscillated plasma reformer was designed to enhance the idea of the

plasma reformer. Its tar destruction performance was achieved for light aromatic tar (i.e.,

benzene). To identify the characteristics of the influential parameters of tar decomposition,

tests were performed by changing the oscillation frequency, the oscillation amplitude, the

steam feed rate, and the total gas feed rate. The tar removal and energy efficiencies were

90.7% and 22.95 g/kWh, respectively. Without oscillation, the decomposition efficiency was

82.6%, and the energy yield was 20.9%, both of which were 8.9% lower than those with the

external oscillation. The test results showed that the EOPR can efficiently destruct tar using

less energy than that used in the gliding arc plasma reformer.

▋ Chapter 6, the combined carbonization-activator was designed for the conversion of

sewage sludge into energy and resources. And the novel thermal treatment system was

suggested as a sequential in-line connection for producing high quality sludge char and

producer gas. A research approach was achieved in 2 steps in the pilot scale test rig for

practical system design.

First, to determine the optimal design conditions, parametric investigations were conducted

on the steam feed rate, the activator temperature and the sludge moisture content. Micro-pores

were well-developed with the steam feed, having best condition of 10 mL/min. However, the

higher heating value of the producer gas decreased with increasing the steam feed rate. And

the pore development in the sludge char improved with increasing the activator temperature.

The higher heating value of the producer gas also increased. The heating value of the

producer gas decreased with increasing the sludge moisture content. The largest micro-pore

development in the sludge char displayed at certain moisture content. Through the parametric

study, the optimal conditions were found to be the steam feed rate of 10 mL/min, activator

temperature of 820℃ and the sludge moisture content of 10.4%.

Second, the integrated thermal system with an in-line connection of the combined

carbonization-activator, the gliding are plasma reformer, and the fixed bed tar adsorber was

developed.

In the carbonization-activator, sludge char and producer gas were produced along with a small

amount of tar. To improve tar adsorption capability of the sludge char, the carbonization-

activator was designed for achieving sequential carbonization and activation. In addition, for

higher producer gas yield and tar destruction, the plasma reformer was installed at the rear

section of the carbonization-activator. The fixed bed adsorber packed with the sludge char

obtained from the carbonization-activator, was tested for adsorption of residual tars.

For the carbonization-activation, the specific surface area of the sludge char was 98.1 m2/g,

with a mean pore size and pore volume of 6.35 nm and 0.2354 cm3/g, respectively. The

producer gases were H2 (41.2%), CO (17.3%), CH4 (9.5%) and CO2 (15.4%). The higher

heating value of the producer gas was 13,400 kJ/Nm3. The gravimetric tar was 26.3 g/Nm

3,

and the total amount of light tar was 10.9 g/Nm3, which contained benzene, naphthalene,

benzonitrile, benzeneacetonitrile, anthracene and pyrene according to the concentration level.

The plasma reformer featured tar cracking and steam reformation, and the decomposition

efficiency of the gravimetric tar was 83.2%, which corresponded to 4.4 g/Nm3 of the

concentration of the gravimetric tar. For light tar, the total amount was 1.3 g/Nm3, which

represented 87.9% of the decomposition efficiency. H2, CO, and CH4 among the components

of the reforming gas were increased, having 13,992 kJ/Nm3 of the higher heating value.

140

The gravimetric tar at the adsorber outlet was 0.5 g/Nm3 with 88.6% of decomposition

efficiency. The total amount of light tar was 0.39 g/Nm3, and the decomposition efficiency

was 40.5%. At the exit of the tar adsorber, 50.5% of H2, 21.9% of CO, 10.5% of CH4, 7.7% of

CO2, and 0.1% of C2H2 were achieved with the higher heating value of 13,482 kJ/Nm3.

Therefore, for the integrated thermal system the carbonization-activation of sewage sludge

can form the sludge char that could be utilized for tar adsorption, and the clean producer gas

was proved to be applicable for gas engines, compressors, etc.

Third, the process analysis for the sequential in-line treatment system was conducted, and the

mass and energy balance was calculated in the each component. The sequential carbonization-

activation system produced the clean gas fuel as useful energy of 40.9% (i.e., net cold gas

efficiency was 40.9%). As the bypass tar should be used as the fuel energy, the system

produced the energy of 41.5% (i.e., the hot gas efficiency was 41.5%). Furthermore, for the

improvement of the system efficiency, two scenarios were suggested, and the verification of

the system efficiency improvement was performed, showing the best process with the cold

gas efficiency of 66.5% and the hot gas efficiency of 67.5%.

141

References

[1] Fytili, D. and A. Zabaniotou (2008). “Utilization of sewage sludge in EU application of

old and new methods—A review.” Renewable and Sustainable Energy Reviews, 12: 116-

140.

[2] Tchobanoglous, G., F. L. Burton, et al. (1991). “Wastewater engineering: treatment,

disposal, and reuse.” New York, USA. McGraw-Hill.

[3] Sands, P. and P. Galizzi (2006). “Council Directive 91/271/EEC of 21 May 1991

concerning urban waste water treatment (OJ L 135 30.05.1991 p. 40)” Documents in

European Community Environmental Law, Cambridge University Press.

[4] Fullana, A., J. A. Conesa, et al. (2003). "Pyrolysis of sewage sludge: nitrogenated

compounds and pretreatment effects." Journal of Analytical and Applied Pyrolysis, 68-69:

561-575.

[5] U.S. Patent, “NITROGEN AND PHOSPHOURS REMOVAL FROM WASTEWATER.”

Assignee: Air Products and Chemicals, Inc., Allentown, Pa., Patent Number: 4,522,722,

Date of Patent: Jun. 11, 1985

[6] Furness, D. T., L. A. Hoggett, et al. (2000). "Thermochemical treatment of sewage

sludge." CIWEM,14: 57-65.

[7] Kasakura, T. and M. Hiraoka (1982). "Pilot plant study on sewage sludge pyrolysis-II."

Water Research, 16(12): 1569-1575.

[8] M. F. Lewis, “Sludge pyrolysis for energy recovery and pollution control.” In Proc. Nat.

Conf. on Municipal Sludge Management and Disposal. Anaheim, California, USA. Aug.

1975. p146.

[9] Bridgwater, A. V. (1989). "Pyrolysis, gasification and liquefaction technologies." G.L.

Ferrero, K. Maniatis, A. Buekens, A.V. Bridgwater (Eds.), Pyrolysis and Gasification,

Elsevier, Amsterdam: 195-198.

[10] Kojima, N., A., Mitomo, et al. (2002). “Adsorption removal of pollutants by active cokes

prodiced from sludge in the energy recycle process of wastes.” Waste Management, 22:

399-404.

[11] Devi, L., K. J. Ptasinski, et al. (2003). "A review of the primary measures for tar

elimination in biomass gasification processes." Biomass and Bioenergy, 24(2): 125-140.

[12] Devi, L., K. J. Ptasinski, et al. (2005). "Pretreated olivine as tar removal catalyst for

biomass gasifiers: investigation using naphthalene as model biomass tar." Fuel

Processing Technology, 86(6): 707-730.

[13] Li, C., D. Hirabayashi, et al. (2009). "Development of new nickel based catalyst for

biomass tar steam reforming producing H2-rich syngas." Fuel Processing Technology,

90(6): 790-796.

[14] Evans, R. J. and T. A. Milne (1987). “Molecular Characterization of the Pyrolysis of

Boimass. 1. Fundamentals.” Energy and Fuels, 1(2): 123-137.

[15] Milne, T. A., N. Abatzoglou, et al. (1998). “Biomass gasifier "tars": Their nature,

formation, and conversion.” National Renewable Energy Laboratory, NREL/TP-570-

25357.

[16] Hasler, P. and T. Nussbaumer (2000). “Sampling and analysis of particles and tars from

biomass gasifiers.” Biomass and Bioenergy, 18: 61-66.

[17] Morf, P., P. Hasler, et al. (2002). “Mechanisms and kinetics of homogeneous secondary

reactions of tar from continuous pyrolysis of wood ships.” Fuel, 81: 843-853.

[18] El-Rub, Z. A. (2008). “Biomass Char as an In-Situ Catalyst for Tar Removal in

Gasification systems.” PhD thesis.

142

[19] Corella, J., A. Orio, et al. (1998). “Biomass Gasification with Air in Fluidized Bed:

Reforming of the Gas Composition with Commercial Steam Reforming Catalysts.”

Industrial & Engineering Chemistry Research, 37(12): 4617-4624.

[20] Elliott, D. C. (1987). “Analysis of medium-BTU gasification condensates.” PNL-5979

UC-61D.

[21] Evans, R. J. and T. A. Milne (1987). “Molecular Characterization of the Pyrolysis of

Biomass. 2. Applications.” Energy and Fuels, 1(4): 311-319.

[22] Hasler, P., R. Bühler, et al. (1997). “Evaluation of Gas Cleaning Technologies for Small

Scale Biomass Gasifiers.” Swiss Federal Office of Energy, Berne.

[23] Overend, R. P., T. A. Milne, et al. (1985). “Fundamentals of thermochemical biomass

conversion.” Elsevier Applied Science Publishers.

[24] Dogru, M., C. R. Howarth, et al. (2002). “Gasification of Hazelnut Shells in a Downdraft

Gasifier.” Energy, 27: 415-427.

[25] Delgado, J., P. M. Aznar, et al. (1996). “Calcined Dolomite, Magnesite, and Calcite for

Cleaning Hot gas from a Fluidized Bed Gasifier with Steam: Life and Usefulness.” Ind.

Eng. Chem. Res., 35: 3637-3643.

[26] Tippayawong, N. and P. Inthasan (2010). “Investigation of Light Tar Cracking in a

Gliding Arc Plasma System.” International Journal of Chemical Reactor Engineering, 8:

1-14.

[27] Chang, J. S. (2001). “Recent Development of Plasma Pollution Control Technology: a

Critical Review.” Science and Technology of Advanced Materials, 2: 571-576.

[28] Yamamoto, T. and S. Futamura (1998). “Non-thermal Plasma Processing for Controlling

Volatile Organic Compounds.” Combustion Science and Technology, 133: 117-133.

[29] Kogelschatz, U. (2004). “Atmospheric Pressure Plasma Technology.” Plasma Physics

and Controlled Fusion, 46: B63-B75.

[30] Paulmier, T. and L. Fulcheri (2005). “Use of Non-thermal Plasma for Hydrocarbon

Reforming.” Chemical Engineering Journal, 106: 59-71.

[31] Petitpas, G., J. D. Rollier, et al. (2007). “A Comparative Study of Non-thermal Plasma

Assisted Reforming Technologies.” International Journal of Hydrogen Energy, 32(14):

2848-2867.

[32] Fridman, A., A. Petrousov, et al. (1994). “Modèle Physique de L'arc Glissant.” Journal de

Physique III France, 4: 1449-1466.

[33] Fridman, A., S. Nester, et al. (1999). “Gliding Arc Gas Discharge.” Progress in Energy

and Combustion Science, 25: 211-231.

[34] Lin, L., B. Wu, et al. (2006). “Characteristics of Gliding Arc Discharge Plasma.” Plasma

Science and Technology, 8(6): 653-655.

[35] Indarto, A., D. R. Yang, et al. (2007). “Advanced VOCs Decomposition Method by

Gliding Arc Plasma.” Chemical Engineering Journal, 131: 337-341.

[36] Kim, H. (2004). “Nonthermal Plasma Processing for Air-pollution Control: a Historical

Review, Current Issues, and Future Prospects.” Plasma Processes and Polymers, 1: 91-

110.

[37] Burlica, R. and B. R. Locke (2008). “Pulsed Plasma Gliding-arc Discharges with Water

Spray.” IEEE Transactions on Industry Applications, 44: 482-489.

[38] Bo, Z., J. Yan, et al. (2007). “The Dependence of Gliding Arc Gas Discharge

Characteristics on Reactor Geometrical Configuration.” Plasma Chemistry and Plasma

Process, 27: 691-700.

[39] Bo, Z., J. Yan, et al. (2004). “Simultaneous Removal of Ethyl Acetate, Benzene and

Toluene with Gliding Arc Gas Discharge.” Journal of Zhejiang University Science A, 9:

143

695-701.

[40] Pemen, A. J. M., S. V. B. van Paasen et al. (2002). “Conditioning of Biomass Derived

Fuel Gas using Plasma Techniques.” 12th European Conference on Biomass for Energy,

Industry and Climate Protection, June 17-21, 2002, Amsterdam, the Netherlands.

[41] Werther, J. and T. Ogada (1999). “Sewage sludge combustion.” Progr Energ Combust

Sci, 25: 55-116.

[42] Song, B. H. (2005). “Gasification Kinetics of Waste Tire Char and Sewage Sludge Char

with Steam in a Thermobalance Reactor.” Ind Eng Chem, 11: 361-367.

[43] Mtsuo, T., T. Hiroshi, et al. (2011). “Modeling of Sludge Behavior in a Steam Dryer.”

Dry Technol, 29: 1748-1757.

[44] Wen, Y. D., H. Y. Jian, et al. (2009). “Measurement and simulation of the contact drying

of sewage sludge in a Nara-type paddle dryer.” Chem Eng Sci, 64(24): 5117-5124.

[45] Arlabosse, P., S. Chavez, et al. (2004). “Method for thermal design of paddle dryers:

Application to municipal sewage sludge.” Dry Technol, 22: 2375-2393.

[46] Arlabosse, P. and T. Chitu (2007). “Identification of the limiting mechanism in contact

drying of agitated sewage sludge.” Dry Technol, 25: 557-567.

[47] Ragnarsson, A. T. (2005). “Sludge Dryer.” United States Patent US 6,892,471 B2.

[48] Bae, H. R. and S. A. Ha (2005). “A Study on the Thermal Kinetic for Disposal of

Sewage Sludge using Thermal Wind Drying of One Mechanical Power (in Korean).” J

of KORRA, 13: 74-84.

[49] Kim, H. S., M. S. Shin, et al. (2005). “A Study for the Thermal Treatment of Dehydrated

Sewage Sludge with Gas-Agitated Double Screw Type Dryer.” J Environ Sci Health A,

40: 203-213.

[50] Dewil, R., J. Baeyens, et al. (2005). “Fenton peroxidation improves the drying

performance of waste activated sludge.” J Hazard Mater, B117: 161-170.

[51] LG Construction Company report (2000) Resources Facility of Sewage Sludge;

ROTADISC system (in Korean). J of KOWREC 8:49-59.

[52] Chun, Y. N., H. W. Song, et al. (2006). “Numerical Simulation of Thermal Flow for

Design of Sludge Dryer (in Korean).” KSWM 2006 Autumn Conference, Suwon, Korea,

397-400.

[53] Deng, W. Y., J. H. Yan, et al. (2009). “Emission characteristics of volatile compounds

during sludges drying process.” J Hazard Mater, 162: 186-192.

[54] Vaxelaire, J. and P. Cézac (2004). “Moisture distribution in activated sludges: a review.

“ Water Research, 38(9): 2215-2230.

[55] Tsang, K. R. and P. A. Vesilind (1990). “Moisture distribution in sludges.” Water

Science and Technology, 22(12): 135-142.

[56] Chun, W. P. and K. W. Lee (2004). “Sludge drying characteristics on combined system

of contact dryer and fluidized bed dryer. Proceedings of the 14th International Drying

Symposium (IDS 2004) São Paulo, Brazil: 1055-1061.

[57] Deng, W. Y., J. H. Yan, et al. (2009). “Measurement and simulation of the contact drying

of sewage sludge in a Nara-type paddle dryer.” Chemical Engineering Science, 64(24):

5117-5124.

[58] Hrischfield, W. A. (1968). “Screw Feeder for Feeding Sewage Sludge or the Like to a

Combustion Chamber.” United States Patent US 3,399,637.

[59] Li, S., S. Xu, et al. (2004). “Fast pyrolysis of biomass in free-fall reactor for hydrogen-

rich gas.” Fuel Process Tech, 85: 1201-1211.

[60] Gonzalez, J. F., S. Roman, et al. (2009). “Pyrolysis of various biomass residues and char

utilization for the production of activated carbons.” J Anal Appl Pyrolysis, 85: 134-141.

144

[61] Zhai, Y. B., Q. Liu, et al. (2008). “Experimental Study on the Characteristics of Sewage

Sludge Pyrolysis Under The Low Temperature Conditions.” Environ Eng Sci, 25: 1203-

1211.

[62] Inguanzo, M., A. Domguez, et al. (2002). “On the pyrolysis of sewage sludge: the

influence of pyrolysis conditions on solid, liquid and gas fractions.” J Anal Appl

Pyrolysis, 63: 209-222.

[63] Lu, C. Q. (1995). “Effect of pre-drying on the pore structure development of sewage

sludge during pyrolysis.” Environ Technol, 16: 495-499.

[64] Ren, L. H., Y. F. Nie, et al. (2006). “Impact of hydrothermal process on the nutrient

ingredients of restaurant garbage.” J Environ Sci, 18: 1012-1019.

[65] Shao, J., R. Yan, et al. (2008). “Pyrolysis characteristics and kinetics of sewage sludge

by thermogravimetry fourier transform infrared analysis.” Energy Fuels, 22: 38-45.

[66] Devki Energy Consultancy Pvt. Ltd., Dryers, 405, Ivory Terrace, R.C. Dutt Road,

Vadodara – 390007. 2006

[67] Arlabosse, P., S. Chavez, et al. (2005). “Drying of municipal sewage sludge: from a

laboratory scale batch indirect dryer to the paddle dryer.” Braz. J. Chem. Eng., 22(2).

São Paulo Apr. 2005

[68] Chae, J. S., Chun, Y. N, Chae, J. W. (1997). “Technology of incineration for low calorific

gas waste.” Journal of Korea Solid Waste Engineering Society, 14(1): 123-131.

[69] Kim, S. C. and Y. N. Chun (2008). “Decomposition characterist of toluene using a

glidarc water-jet plasma.” Journal of Korean Society for Atmospheric Environment,

24(3): 329-335.

[70] Kim, S. C., H. W. Song, Y. N. Chun (2006). “Emission characteristic for high efficiency

and low NOx of externally oscillated burner.” Journal of Korean Society for

Atmospheric Environment, 22(5): 693-700.

[71] Abrego, J., J. s. Arauzo, et al. (2009). "Structural Changes of Sewage Sludge Char during

Fixed-Bed Pyrolysis." Industrial & Engineering Chemistry Research, 48(6): 3211-3221.

[72] Karayildirim, T., J. Yanik, et al. (2006). "Characterisation of products from pyrolysis of

waste sludges." Fuel, 85(10-11): 1498-1508.

[73] Nipattummakul, N., I. I. Ahmed, et al. (2010). "Hydrogen and syngas production from

sewage sludge via steam gasification." International Journal of Hydrogen Energy,

35(21): 11738-11745.

[74] Midilli, A., M. Dogru, et al. (2002). "Hydrogen production from sewage sludge via a

fixed bed gasifier product gas." International Journal of Hydrogen Energy, 27(10):

1035-1041.

[75] Park, H. J., H. S. Heo, et al. "Clean bio-oil production from fast pyrolysis of sewage

sludge: Effects of reaction conditions and metal oxide catalysts." Bioresource

Technology, 101(1, Supplement 1): S83-S85.

[76] Fonts, I., M. Azuara, et al. (2009). "Study of the pyrolysis liquids obtained from different

sewage sludge." Journal of Analytical and Applied Pyrolysis, 85(1-2): 184-191.

[77] Bandosz, T. J. and K. Block (2006). "Effect of pyrolysis temperature and time on

catalytic performance of sewage sludge/industrial sludge-based composite adsorbents."

Applied Catalysis B: Environmental, 67(1-2): 77-85.

[78] Martin, M. J., E. Serra, et al. (2004). "Carbonaceous adsorbents from sewage sludge and

their application in a combined activated sludge-powdered activated carbon (AS-PAC)

treatment." Carbon, 42(7): 1389-1394.

[79] Sutcu, H. (2008). "Pyrolysis of PhragmitesAustralis and characterization of liquid and

solid products." Journal of Industrial and Engineering Chemistry, 14(5): 573-577.

145

[80] Fonts, I., E. Kuoppala, et al. (2009). "Physicochemical Properties of Product Liquid from

Pyrolysis of Sewage Sludge." Energy Fuels, 23(8): 4121-4128.

[81] Phuphuakrat, T., T. Namioka, et al. (2010). "Tar removal from biomass pyrolysis gas in

two-step function of decomposition and adsorption." Applied Energy, 87(7): 2203-2211.

[82] Good, J., L. Ventress, et al. (2005). "Sampling and analysis of tar and particles in

biomass producer gases." Technical Report, BTG biomass technology group CEN

BT/TF 143.

[83] Yamazaki, T., H. Kozu, et al. (2005). "Effect of Superficial Velocity on Tar from

Downdraft Gasification of Biomass." Energy & Fuels, 19(3): 1186-1191.

[84] Neeft, J. P. A. (2005). "Rationale for setup of inpinger train." SenterNovem CEN BT/TF

143: 1-14.

[85] Son, Y. I., M. Sato, et al. (2009). "A Study on Measurement of Light Tar Content in the

Fuel Gas Produced in Small-Scale Gasification and Power Generation Systems for Solid

Wastes." Journal of Environment and Engineering, 4(1): 12-23.

[86] Hosoya, T., H. Kawamoto, et al. (2008). "Pyrolysis gasification reactivities of primary tar

and char fractions from cellulose and lignin as studied with a closed ampoule reactor."

Journal of Analytical and Applied Pyrolysis, 83(1): 71-77.

[87] Umeki, K. (2009). "Modelling and simulation of biomass gasification with high

temperature steam in an updraft fixed-bed gasifier." Doctoral thesis, Tokyo Institute of

Technology.

[88] Mašek, O., M. Konno, et al. (2008). "A study on pyrolytic gasification of coffe grounds

and implication to allothermal gasification." Biomass and Bioenergy, 32: 78-89.

[89] Zhang, B., S. Xiong, et al. (2011). “Mechanism of wet sewage sludge pyrolysis in a

tubular furnace.” Int. J. Hydrogen Energy, 36: 355-363.

[90] Pütün, E., F. Ates, et al. (2008). “Catalytic pyrolysis of biomass in inert and steam

atmospheres.” Fuel, 87: 815-824.

[91] Chent, H., T. Namiokal, et al. (2010). “Comparison of Tar Behavior and Characteristics

during Pyrolysis and Reforming of Sewage Sludge and Wood Chips.” 8th International

SymposiullI on High Temperature Air Combustion and Gasification, Poznan, Poland:

July, 5 -7: 445-454

[92] IUPAC (1982). "Manual of Symbols and Terminology of Colloid Surface." Butterworths,

London.

[93] Bagreev, A., T. J. Bandosz, et al. (2001). "Pore structure and surface chemistry of

adsorbents obtained by pyrolysis of sewage sludge-derived fertilizer." Carbon, 39(13):

1971-1979.

[94] Inguanzo, M., J. A. Mendez, et al. (2001). "Reactivity of pyrolyzed sewage sludge in air

and CO2." Journal of Analytical and Applied Pyrolysis, 58-59: 943-954.

[95] Khalili, N. R., M. Campbell, et al. (2000). "Production of micro- and mesoporous

activated carbon from paper mill sludge: I. Effect of zinc chloride activation." Carbon,

38(14): 1905-1915.

[96] Sinfelt, J. H. and J. C. Rohrer (1962). "CRACKING OF HYDROCARBONS OVER A

PROMOTED ALUMINA CATALYST." The Journal of Physical Chemistry, 66(8):

1559-1560.

[97] Yaman, S. (2004). "Pyrolysis of biomass to produce fuels and chemical feedstocks."

Energy Conversion and Management, 45(5): 651-671.

[98] James, R., W. Couper, et al. “Fair & Stanley M. Wales, Chemical Process Equipment-

Selection & Design, CHAPTER 17 CHEMICAL REACTORS, Revised 2nd Edition,

Butterworth-Heinemann ISBN: 9780123725066.

http://www.download-it.org/learning-resources.php?promoCode=&partnerID=&content=partnerlist&partnerIDlook=3

146

[99] Lillo-Ródenas, M. A., A. J. Fletcher, et al. (2006).“Competitive adsorption of a benzene–

toluene mixture on activated carbonsat low concentration.” Carbon, 44:1455–1463.

[100] Zhao, H., J. Dirk, et al. (2000). "Performance of a Nickel-Activated Candle Filter for

Naphthalene Cracking in Synthetic Biomass Gasification Gas." Industrial &

Engineering Chemistry Research, 39(9): 3195-3201.

[101] Piroonlerkgul,P., S.Assabumrungrat, et al. (2008). "Selection of appropriate fuel

processor for biogas-fuelled SOFC system." Chemical Engineering Journal, 140(1-3):

341-351.

[102] Kolbitsch, P., C. Pfeifer, et al. (2008). "Catalytic steam reforming of model biogas."

Fuel, 87(6): 701-706.

[103] Rasi, S., A. Veijanen, et al. (2007). "Trace compounds of biogas from different biogas

production plants." Energy, 32(8): 1375-1380.

[104] Chun, Y. N. and H. O. Song (2008). "Syngas Production Using Gliding Arc Plasma."

Energy Sources, Part A: Recovery, Utilization, and Environmental Effects, 30(13):

1202-1212.

[105] Indarto, A., J. W. Choi, et al. (2006). "Effect of additive gases on methane conversion

using gliding arc discharge." Energy, 31(14): 2986-2995.

[106] Sreethawong, T., P. Thakonpatthanakun, et al. (2007). "Partial oxidation of methane

with air for synthesis gas production in a multistage gliding arc discharge system."

International Journal of Hydrogen Energy, 32(8): 1067-1079.

[107] Fridman, A., S. Nester, et al. (1998). "Gliding arc gas discharge." Progress in Energy

and Combustion Science, 25(2): 211-231.

[108] Indarto, A., D. R. Yang, et al. (2007). "Gliding arc plasma processing of CO2

conversion." Journal of Hazardous Materials, 146(1-2): 309-315.

[109] Yu, L., X. Tu, et al. (2010). "Destruction of acenaphthene, fluorene, anthracene and

pyrene by a dc gliding arc plasma reactor." Journal of Hazardous Materials, 180(1-3):

449-455.

[110] Kabashima, H., H. Einaga, et al. (2003). "Hydrogen generation from water, methane,

and methanol with nonthermal plasma." Industry Applications, IEEE Transactions on,

39(2): 340-345.

[111] Lee, D. H., K. T. Kim, et al. (2007). "Optimization scheme of a rotating gliding arc

reactor for partial oxidation of methane." Proceedings of the Combustion Institute, 31(2):

3343-3351.

[112] Zhang, Z. G., G. Xu, et al. (2004). "Process development of hydrogenous gas

production for PEFC from biogas." Fuel Processing Technology, 85(8-10): 1213-1229.

[113] Effendi, A., K. Hellgardt, et al. (2005). “Optimising H2 production from model biogas

via combined steam reforming and CO shift reactions.” Fuel, 84: 869-874.

[114] Hammer, T., T. Kappes, et al. (2003). “Plasma Catalytic Hybrid Reforming of Methane.

Utilization of Greenhouse Gases”, American Chemical Society. 852: 292-301.

[115] Devi, L., K. J. Ptasinski, et al. (2005). "Decomposition of Naphthalene as a Biomass

Tar over Pretreated Olivine: Effect of Gas Composition, Kinetic Approach, and

Reaction Scheme." Industrial & Engineering Chemistry Research, 44: 9096-9104.

[116] Chun, Y. N. and H. O. Song (2006). "Syngas Production from Propane Using Gliding

Arc Plasma Reforming." ENVIRONMENTAL ENGINEERING SCIENCE, 23(6):

1017-1023.

[117] Pemen, A. J. M., S. A. Nair, et al. (2003). "Pulsed Corona Discharges for Tar Removal

from Biomass Derived Fuel Gas." Plasmas and Polymers, 8(3): 209-224.

147

[118] Wu, Z. L., X. Gao, et al. (2004). "Decomposition characteristics of toluene by a corona

radical shower system." Journal of environmental sciences, 16(4): 543-547.

[119] Guo, Y. F., D. Q. Yea, et al. (2006). "Toluene decomposition using a wire-plate

dielectric barrier discharge reactor with manganese oxide catalyst in situ." Journal of

Molecular Catalysis A: Chemical, 245: 93-100.

[120] Du, C. M. and J. H. Yan (2007). "Decomposition of toluene in a gliding arc discharge

plasma reactor." Plasma Sources Science. Technology 16(4): 791-797 .

[121] Indarto, A. and J. W. Choi (2006). "Effect of additive gases on methane conversion

using gliding arc discharge." Energy, 31(14): 1067-1079.

[122] Korean Standards Association, KS I 2211 (2006). “Methods for Determination of

benzene in flue gas.”

[123] Yu, L., X. Li, et al. (2009). “Decomposition of Naphthalene by dc Gliding Arc Gas

Discharge.” J. Phys. Chem. A., 114: 360-368.

[124] Du, C., J. Yan, et al. (2006). "Simultaneous Removal of Polycyclic Aromatic

Hydrocarbons and Soot Particles from flue Gas by Gliding arc Discharge Treatment."

Plasma Chemistry and Plasma Processing, 26(5): 517-525.

[125] Dayton, D. (2002). “A Review of the Literature on Catalytic Biomass Tar Destruction.”

National Renewable Energy Laboratory NREL/TP-510-32815, 1-27.

[126] Chun, Y. N., Y. C. Yang, et al. (2009). “Hydrogen generation from biogas reforming

using a gliding arc plasma-catalyst reformer.” Catalysis Today, 148(3-4): 283-289.

[127] Benstaali, B., P. Boubert, et al. (2002). “Density and Rotational Temperature

Measurements of the OH and NO Radicals Produced by a Gliding Arc in Humid Air.”

Plasma Chem. Plasma Process., 22: 553-571.

[128] Bo, Z., J. Ya, et al. (2008). “Scale-up analysis and development of gliding arc discharge

facility for volatile organic compounds decomposition.” J. Hazard. Mater., 155: 494-

501.

[129] Shiki, H., T. Okawa, et al. (2008). “Development of split gliding arc for surface

treatment of conductive material.“ Thin Solid Films., 516: 3684-3689.

[130] Guo, Y., X. Liao, et al. (2008). "Detection of hydroxyl radical in plasma reaction on

toluene removal." Journal of Environmental Sciences, 20(12): 1429-1432.

[131] Sakuntala, M. and V. K. Jain (1978). “Acoustic wave interaction with plasma.” J. Phys.

D: Appl. Phys., 11: 1925-1929.

[132] Giacomazzi, E., D. Cecere, et al. (2009). “Effects of Forced Acoustic Waves onto Jet

Shear Layers.” Combustion Colloquia 32nd Combustion Meeting Napoli, 26 April 2009.

[133] M. Okada, T. Nakane, et al. (2009). “Effect of Sound Wave Irradiation on Methane

Conversion in DC Pulse Discharge Plasma.” Chem. Prod. Process Model., 4(5): 1-10.

[134] Okada, M., T. Nakane, et al. (2008). “Effects of Sound-wave Irradiation on

Decomposition of Carbon Dioxide in DC-pulse Discharge Field.” Journal of the Japan

Petroleum Institute, 51(3): 180-185.

[135] Domínguez, A., J. A. Menéndez, et al. (2006). “Hydrogen rich fuel gas production from

the pyrolysis of wet sewage sludge at high temperature.” Journal of Analytical and

Applied Pyrolysis, 77 (2): 127-132.

[136] Dogru, M., A. Midilli, et al. (2002). “Gasification of sewage sludge using a throated

downdraft gasifier and uncertainty analysis.” Fuel Processing Technology, 75 (1): 55-82.

[137] Phuphuakrat, T., N. Nipattummakul, et al. (2010). “Characterization of tar content in the

syngas produced in a downdraft type fixed bed gasification system from dried sewage

sludge.” Fuel, 89 (9): 2278-2284.

148

[138] Devi, L., K. J. Ptasinski, et al. (2005) “Catalytic decomposition of biomass tars: use of

dolomite and untreated olivine.” Renewable Energy, 30 (4): 565-587.

[139] Bergman, P. C. A., S. V. B. van Paasen, et al. (2002). “The novel “OLGA” technology

for complete tar removal from biomass producer gas, Pyrolysis and Gasification of

Biomass and Waste.” Expert Meeting, Strasbourg, France, 1-15.

[140] Zhang, K., H. T. Li, et al. (2009). “The thermal cracking experiment research of tar

model compound.” International Conference on Energy and Environment Technology,

1-5.

[141] Pfeifer, C. and H. Hofbauer (2008). “Development of catalytic tar decomposition

downstream from a dual fluidized bed biomass steam gasifier.” Powder Technology,

180 (1-2): 9-16.

[142] Hosokai, S., J. I. Hayashi, et al. (2005). “Spontaneous generation of tar decomposition

promoter in a biomass steam reformer.” Chemical Engineering Research & Design, 83

(A9): 1093-1102.

[143] Onozaki, M., K. Watanabe, et al. (2006), “Hydrogen production by the partial oxidation

and steam reforming of tar from hot coke oven gas.” Fuel, 85 (2): 143-149.

[144] Nair, S. A., A. J. M. Pemen, et al. (2003). “Tar removal from biomass-derived fuel gas

by pulsed corona discharges.” Fuel Processing Technology, 84 (1-3): 161-173.

[145] Nair, S. A., K. Yan, et al. (2005). “Tar Removal from Biomass Derived Fuel Gas by

Pulsed Corona Discharges: Chemical Kinetic Study II.” Industrial & Engineering

Chemistry Research, 44 (6): 1734-1741.

[146] Du, C. M., J. H. Yan, et al. (2007). “Decomposition of toluene in a gliding arc discharge

plasma reactor.” PLASMA SOURCES SCIENCE AND TECHNOLOGY 16: 791-797.

[147] IUPAC (1972). “Manual of Symbols and Terminology, Appendix 2, Part 1, Colloid and

Surface Chemistry.” Pure Appl. Chem., 31: 578.

[148] Xiao, R., X. Chen, et al. (2010). “Pyrolysis pretreatment of biomass for entrained-flow

gasification.” Applied Energy, 87(1): 149-155.

[149] Dominguez, A., J. A. Menendez, et al. (2006). “Hydrogen rich fuel gas production from

the pyrolysis of wet sewage sludge at high temperature.” J Anal Appl Pyrol, 77: 127-

132.

[150] Chun, Y. N. (2002). “Incineration and Air Pollution Control.” Chosun University Press.

[151] ArlabosseI, P., S. ChavezI, et al. (2005). “Drying of municipal sewage sludge: from a

laboratory scale batch indirect dryer to the paddle dryer.” Braz. J. Chem. Eng., 22(2):

227-232.

[152] Thunman, H., Niklasson, F., et al. (2001). “Composition of Volatile Gases and

Thermochemical Properties of Wood for Modeling of Fixed or Fluidized Beds.” Energy

& Fuels, 15: 1488-1497.

[153] Park, Y. T. (2007). “Activated carbon technology.” Dongwha Tech, Press.

ACKNOWLEDGMENT This dissertation was sponsored by JSPS (Japan Society for the Promotion

of Science) as a JSPS RONPAKU（Dissertation PhD）Program (ID No.

KOSEF-10814). Author is thankful for the support of JSPS fellowship.