pa ethan om 2012

Fuel Processing Technology 104 (2012) 144–154

Contents lists available at SciVerse ScienceDirect

Fuel Processing Technology

j ourna l homepage: www.e lsev ie r .com/ locate / fuproc

Performance of tar removal by absorption and adsorption for biomass gasification

Anchan Paethanom ⁎, Shota Nakahara, Masataka Kobayashi, Pandji Prawisudha, Kunio YoshikawaDepartment of Environmental Science and Technology, Tokyo Institute of Technology, G5-8, 4259 Nagatsuta, Midori-ku, Yokohama, Kanagawa 226–8502, Japan

⁎ Corresponding author. Tel.: +81 45 924 5507; fax:E-mail address: [email protected] (A. P

0378-3820/$ – see front matter © 2012 Elsevier B.V. Alldoi:10.1016/j.fuproc.2012.05.006

a b s t r a c t
a r t i c l e i n f o
Article history:Received 26 January 2012Received in revised form 1 May 2012Accepted 3 May 2012Available online 27 May 2012

Keywords:Biomass gasificationTar removalAbsorptionAdsorptionGas cleaning

Biomass gasification is an attractive and successful waste-to-energy technology. Even though it has been per-forming effectively, many troubles are still occurring. For advanced applications, gas needs to be clean enoughand tar should extensively be removed. Otherwise, tar in the producer gaswill condense at reduced temperatureandwill cause blocking and fouling of engines. Physical tar removal is proven to be technically and economicallyattractive approach for gas cleaning. In this paper, three tar removal techniques were investigated for each typeof tar; 1) heavy tar removal by absorption using vegetable oil andwaste-cooking oil scrubbers, 2) light tar remov-al by adsorption using rice husk and rice husk char adsorbent bed and 3) heavy tar removal by combination ofabsorption and adsorption using vegetable oil scrubber and rice husk char adsorbent bed. Temperature of thethermal tar decomposition process was set at 800 °C and temperature of the physical gas cleaning unit ccwasat room temperature. The result showed that the absorption technique was effective for heavy tar removaland the adsorption technique was capable of light tar removal. By combining vegetable oil scrubber and ricehusk char adsorbent bed, 95.4% of gravimetric tar could be successfully removed.

© 2012 Elsevier B.V. All rights reserved.

1. Introduction

Compared to the costly and rare petroleum fuel, alternative energyresources like biomass, including agriculture wastes, which are contin-uously increasing with the world population growth and economic de-velopment, are considered as highly efficient renewable energyresources for energy needs throughout the world [1,2]. Energy frombiomass can significantly reduce the green house gas emissions andproblems, which cause global warming and climate change [3].

Nevertheless, even though there aremanywaste-to-energy technol-ogies, including combustion, gasification and pyrolysis, which havelately been successfully demonstrated [4–9], commercialized or underimplementation, still troubles occur during the process and loads of ag-riculture wastes are still abundant. For that reason, waste-to-energytechnologies need to be more studied and researched.

In agricultural countries, lots of agriculture residues or biomasswastes, such as rice husk and woods, are produced in each year [10].The world annual production of rice husk amounts to more than 120million tons [11]. These biomass wastes are one of the main resourcesof renewable energy. Accordingly, biomass gasification is an attractivetechnology for advanced application such as electricity generation andliquid or gaseous transportation fuel production. For these applications,the producer gas needs to be cooled down, de-dusted and tar shouldsignificantly be eliminated. Anyhow, the gasification technology for

+81 45 924 5518.aethanom).

rights reserved.

biomass is still in the development stage. The main technical obstacleis the efficient elimination of tar from the producer gas.

Biomass tar is a complex compound, consisting of hundreds hydro-carbon compositions. Tar could be defined as a complex mixture of or-ganic hydrocarbon compounds, which is benzene and larger thanbenzene [12,13]. Tar can be characterized into different classifications[14,15], in which two main classifications are light tar and heavy tar.Light tar is a mixture of light heterocyclic compounds, such as phenoland naphthalene, some of which are water-soluble. For heavy tar, itwill condense at the reduced gas temperature [16] and cause majorclogging, fouling, efficiency loss and unscheduled plant stops [14].These two main kinds of tar; heavy and light, were both concerned inthis research and the most effective system, which is able to removeboth kind of tar was aimed. Since the nature of heavy tar, normally pre-sents in the droplet form, and light tar, usually in vapor form, differentand individual tar removal techniques had to be investigated.

1.1. Tar removal technique

Many of the tar removal research work have been aimed for devel-opment of advanced techniques that able to remove tar effectivelysuch as using catalytic cracking, thermal cracking, advanced oil scrubbertechnology, or plasma reactor for a range of objective, where most ofthem are costly and not able to apply to some projects such as gasifica-tion for rural electrification. On the contrary, this research is prospectiveto provide fundamental technique that concernmostly on economicallyand simply operable system for tar removal with preservation of effec-tiveness and potential for scaling up. Therefore, waste resources havebeen used as tar removal mediums.

http://dx.doi.org/10.1016/j.fuproc.2012.05.006

mailto:[email protected]

http://dx.doi.org/10.1016/j.fuproc.2012.05.006

http://www.sciencedirect.com/science/journal/03783820

145A. Paethanom et al. / Fuel Processing Technology 104 (2012) 144–154

There are many possible techniques of tar removal depending onwhere tar is removed [17]. In primary methods, tar will be removedinside the combustion or gasification process and for secondarymethods, which is more efficient, economical and easier to controlbecause tar will be removed by outside installed gas cleaning equip-ments. Gas cleaning methods can also be characterized into chemical(catalytic and non-catalytic such as thermal cracking) and physical(such as adsorption and absorption) processes. The physical processis an attractive, technically and economically feasible tar removalmeth-od. Moreover, physical tar removal process is uncomplicated adaptableto any gasification system. However, it also depends on gas qualityspecifications required for specific downstream applications. Physicaltar removal could be characterized as wet system such as spray towersor packed column scrubber and dry systemsuch as cyclone,filters or ad-sorbents [18].

For wet system, many gasification plants commonly utilize varie-ty of water scrubber structure to remove tar and other gas contami-nants such as spray tower, impingement scrubber, packed bedscrubber and venturi scrubber [18–20]. Among all these kind ofstructures, water spray tower is the simplest and economical scrub-ber structure. However, using water as scrubber, subsequent to lotsof unsatisfied drawback especially expensive expense on wastewa-ter treatment and regeneration. Moreover, due to the hydrophobiccharacteristic of water, it has shown low solubility of tar compounds.To conquer the weaknesses of water scrubber and for more effectivetar removal, oil scrubbers have been widely studied [21–25]. Eachtype of oil absorbents was deliberated and they performed as hightar absorption mediums at higher than 60% tar removal comparedto water, which can only remove hydrophilic tar compounds andonly 38.9% hydrophobic tar compounds. The oil scrubber perfor-mance resulted as diesel>vegetable oil>biodiesel>engine oil. Nev-ertheless, the author suggested that vegetable oil should beconsidered as best and most effective adsorbent among others dueto low evaporating point of diesel oil that effected the increasing ofthe gravimetric tar to the downstream application [21]. For stateddemonstration in bigger scale; Güssing Plant, Austria, benzene andremaining tar compounds are greatly removed by rapeseed-methyl-ester (RME) scrubber working at 5 °C [22]. Not only the absorption me-dium itself, but the conceptual design of scrubbing device technologyseems to havemade advancement such as OLGA technology, developedby the Energy research Centre of the Netherlands, using special oil ab-sorber tower with striper tower to regenerate the oil by using air orsteam to deabsorb tar. OLGA shows more than 99% of heavy and lighttar removal performance [14,15]. These advance technologiesmay suit-able for some applications such as fuel cells and chemical synthesis likeFischer–Tropsch. However, concerning high energy required to cooldown the RME scrubber and OLGA high production cost, these ad-vanced tar removal technologies may affect the cost feasibility. Simpleand economical but effective tar removalmethod, which is also applica-ble for some projects such as rural electrification have been the mainconcerns in this research paper.

Because of the expensive price of vegetable oil, the cost of vegeta-ble oil-based gas cleaning unit operation is high. On the other hand,the rising production of used cooking oil from household, restaurantsand industrial sources is causing big environmental problems in theworld [26]. This used cooking oil is commonly poured down to thedrain as a waste, resulting in problems for wastewater treatment oris integrated into the food chain through animal feeds, accordinglybecoming a cause of human health harms. The use of this waste-cooking oil as a scrubber medium for tar removal in biomass gasifica-tion systems should be one of the procedures for solving the twintroubles of environment pollution and energy shortage. This tech-nique has not yet been found investigated or reported beforehandat all. Therefore, in this research, each operating conditions of vegeta-ble oil and waste-cooking oil as scrubbing mediums for tar removalhave been evaluated due to its hydrophilic property and low cost.

Another physical tar removal technique is a dry system, to particu-larly remove tar components in vapor phase. Among various physicaltechniques, adsorption process is considered better as compared toother methods because of convenience, easy operation and simplicityof design. It has been reported that activated carbon performs as an ef-fective adsorbent due to its high porosity property [27–31]. Activatedcarbon can be prepared from variety of carbon containing materials,such as coal or any biomass. In any thermal or gasification processwith biomass as feedstock, biomass char will be produced. For biomasschar, it also has been reported to have removal ability for model tarcomponents [32–34]. Biomass char as tar adsorbent investigation hasbeen found in some publications. Pyrolyzed char from wood samplesover a hot bed has been used for tar reduction. It was found that thepresence of wood char could reduce the amount of tar. The carbonyield in the condensable phases was 37.6 wt.% by pyrolysis at 500 °Cdown to 15.3 wt.% without char in the 300 mm-long tar cracking zoneat 800 °C and further down to 12.9 wt.% by introduction of char in thezone. The heavy condensable phase accounting for 18.4 wt.% of the in-puttedwood at 500 °Cdecreased to about 8 wt.% at 800 °C [35]. Biomasschar could also be active like catalyst. It was compared with other cata-lysts such as calcined dolomite, olivine and nickel for the cracking ofphenol and naphthalene. The results presented that the char was an ef-fective and low cost catalyst for tar removal [32]. Although phenol andnaphthalene are tar components, the real tar generated directly frombiomass should be investigated for more reliable results for scaling up.Further research of char for tar removal was also reported with reactordesign as the TREC-reactor (Tar REmoval by Char) [36].

Subsequently, waste char generated from rice husk gasification is at-tractive to be tar removal medium as also a porous carbon material [34]and now considered as waste needed to be taken care of. Consideringscaling up for industrial process utilization, authentic biomass tar shouldbe utilized for the evaluation of adsorption medium candidates insteadof usingmodel tar components. In addition, most of the tar removal liter-atures concentrate on individual technique for tar removal such as ther-mal cracking, catalytic cracking, filter, scrubber or adsorption mediumand each technique has different noteworthy in itself. Therefore, the hy-brid of multiple tar removal techniques should be investigated.

Normally, for tar adsorption technique, heavy tars in liquid form caneasily accumulate on the filter surface that leads to eventually pluggingand not easy to clean. To get rid of this problem and prolong tar adsor-bent lifetime, the combination of adsorption techniquewith absorptionor wet scrubber should be consider. Thewet scrubber using oil materialshould be installed for heavy tar removal and then the leave light tar invapor form to be adsorbed by char adsorbent unit.

In this paper, authentic biomass tar generated from pyrolysis pro-cess followed by thermal decomposition has been used for evaluatingthree physical tar removal techniques; 1) heavy tar removal by absorp-tion using vegetable oil andwaste-cooking oil scrubberswith the turbu-lent mixing, 2) light tar removal by adsorption using rice husk and ricehusk char adsorbent bed and 3) heavy tar removal by combination ofabsorption and adsorption using vegetable oil scrubber and rice huskchar adsorbent bed. The temperature of the tar thermal decompositionprocess was set at 800 °C and the temperature of physical gas cleaningwas the room temperature. For heavy tar, the results were analyzedand discussed in the term of gravimetric tar removal performancesusing wet tar measurement method and for light tar, the results wereanalyzed using the dry tar measurement method.

2. Materials and experimental setup

2.1. Experimental setup

For this tar removal experiment, an externally heated continuous-type pyrolyzer reactor has been used to generate tar from rice huskfeedstock. The characterizations of rice husk feedstock are shown inTable 1. The process scheme of the experimental setup is shown in

Table 1The proximate and ultimate analysis of rice husk.

wt.% dry basis

Ultimate analysis (ash free basis)C 32.2%H 4.3%N 0.8%Cl 0.4%

Proximate analysisVolatile matter content 59.7%Fixed carbon 11.9%Ash 28.4%

146 A. Paethanom et al. / Fuel Processing Technology 104 (2012) 144–154

Fig. 1. The rice husk feedstock obtained from Thailand was prepared bycrushing and sieving with the mesh size of 0.125–0.500 mm. The feed-stock was dried in an oven at 105 °C overnight for moisture eliminationbefore packing in the feeder. The screw feeder controlled the feed rateat 0.6 g/min continuously. The pyrolyzer reactor was made from stain-less steel with the inner diameter of 30 mm and the length of280 mm, and was surrounded by an electric heater. The pyrolyzer tem-perature was set at 800 °C and kept for 30 min before the start of thefeeder. When the feedstock was supplied into the pyrolyzer, thevolatiles of the feedstock were released as syngas, called producer gas.This tar-contained producer gas with nitrogen as carrier gas enteredinto the reformer, which was also kept at 800 °C in order to conductthermal decomposition of tar. The reformer reactor was also madefrom stainless steel with the inner diameter of 25 mm and the lengthof 1300 mm, and was surrounded by an electric heater. After that, thegas flowed to the gas cleaning unit at the flow rate of 1.5 l/min. The pro-ducer gas was sampled by wet or dry tar measurement method at theexit of the gas cleaning unit. The schematics of each type of gas cleaningunit used in this paper are shown in Fig. 2. As shown in Fig. 1, the gascleaning unit was installed between the reformer and the gas samplingline consisting of wet or dry tar measurement instruments describedlater.

2.2. Absorbents

Two kinds of scrubbing medium used in this research were vegeta-ble oil and waste-cooking oil. Vegetable oil was purchased from J-OILMILLS Ltd. The main ingredients were 60% of soybean oil and 40% of ca-nola oil. Waste-cooking oil was obtained from Best Trading Company,Atsugi, Kanagawa, Japan. The crude waste-cooking oil material wasbought and collected from communities nearby (households, restau-rants, hotels, etc.) then mixed and put in a 200 L tank for pretreatmentprocess. After left in the tank for one day, 5–10 L of oil at the tank bot-tom was removed to get rid of the sediment and contaminate. The

Gas Cleaning Unit

T

T

T

T

T

Nitrogencylinder

To sampling line

T Thermocouple

Fig. 1. The process scheme of

pressure was reduced to 0.095 MPa and kept at 65 °C for water reduc-tion. About 100 ml of water evaporated from 200 L crude waste-cooking oil. The characterizations of vegetable oil and pretreatedwaste-cooking oil used in this experiment were measured at ambienttemperature (27 °C) as shown in Table 2.

2.3. Adsorbents

The rice husk and rice husk char adsorbent used in this research hadbeen sent from Thailand. This rice husk char was carbonized and left asresidue in downdraft gasification process at 500 °C. The scanning electronmicroscope (SEM) micrograph and the proximate and ultimate analysisof rice husk and rice husk char adsorbent along with BET specific surfacearea data are shown in Fig. 3. and Table 3, respectively. From these char-acterizations, it can be seen that after high temperature gasified, rice huskchar became a highly porous material. Therefore, this char was found tohave suitable characteristics to be used as a tar removal adsorbent. Theadsorbents were prepared from crushing and sieving with the meshsize of 0.5–0.1 mm and dried in an oven at 105 °C overnight for moistureelimination.

3. Sampling and analysis method

Biomass tars are complex compounds, consisting of hundreds hy-drocarbon compositions. The tars in this research were sampled andtheir amounts were analyzed by two different methods; wet and dry.These two methods have different prospective and technique. The wetmethod aims to analyze condensed high molecular weight tar dropletsand determined by weight as gravimetric tar, which is called heavy tar.The dry method aims to analyze the volatile tar aerosols in the gas flowand fractionated by GC-FID, which is called light tar. The heavy tar com-ponents are not able to elute through the GC columns. The advantageand disadvantage comparisons between wet and dry tar measurementmethod used in this research are shown in Table 4.

The details of two methods of tar measurement had been used inthis research are as following;

3.1. Wet method for heavy tar measurement

This method has been modified along with the guideline for sam-pling and analysis of tar and particles in biomass producer gases [37].The producer gas passed through a series of 10 impinger bottles filledwith 100 ml isopropanol and tar was collected by both condensationand absorption. This impinger train was connected with filters, a gasflow meter, a manual control valve, and a suction pump, respectivelyas shown in Fig. 4. The entire impinger train was placed inside thecooling baths. The bath temperature for the first 6 impinger bottleswere 3±1 °C with the mixture of salt, water and ice and for the last 4

T

Screw feeder

Pyrolyzer

Reformer

Gas flow control valve

the experimental setup.

T Thermocouple

T

Magnetic stirrer machine

Oil scrubber

Gas Exit Gas In

T

Gas In

Gas Exit

T

Magnetic stirrer machine

Oil scrubber

Gas In

T

Gas Exit

a) b)

c)

Fig. 2. Schematics of each type of gas cleaning unit a) the oil scrubber for the absorption technique, b) the adsorbent bed for the adsorption technique and c) combination of the oilscrubber and the adsorbent bed for the combination of absorption and adsorption.

a)


impinger bottleswere−22±1 °Cwith the isopropanol andmechanicalcooling device. This cooling procedure needed approximately 30 min.The sampled producer gas was controlled at the flow rate of approxi-mately 0.8 l/min for 48 min. After sampling, all of isopropanol samplingsolvent in each impinger bottleweremixed together, filtrated and driedby a standard rotary evaporator in a water bath kept at 40 °C. Then,weigh the flask accurately and determine the amount of residue,which was heavy tar. This heavy tar was defined as gravimetric tar.

b)

3.2. Dry method for light tar measurement

The procedure of dry type sampling method is presented in Fig. 5.A series of 6 mm I.D. charcoal tube containing 150 mg of activatedcarbon and 8 mm I.D. silica gel tube containing 780 mg of silica gelwas connected to collect tar by adsorption and condensation strate-gies, under the room temperature. Both of sampling tubes were pur-chased from Sibata Sciencetific Technology Ltd. with 90.0–97.4%accuracy for aromatic hydrocarbon standards. The charcoal tube

Table 2The characterizations of vegetable oil and waste-cooking oil (at 27 °C).

Parameters Density (g/cm3)

Kinematic Viscosity(mm2/s)

H C N

Vegetable oil 0.84 42.9 10.41 77.38 –

Waste-cookingoil

0.88 53.9 11.63 77.48 0.08Fig. 3. Scanning electron microscope (SEM) micrograph of a) rice husk and b) rice huskchar.

Table 3Proximate and ultimate analysis of rice husk and rice husk char.

Parameters Ultimate analysis(wt.% dry and ashfree basis)

Proximate analysis(wt.% dry basis)

BET Specific surfacearea (m2/g)

C H N Cl Volatilematter

Fixedcarbon

Ash

Rice husk 32.2 4.3 0.8 0.4 59.7 11.9 28.4 2.2Rice huskchar

16.9 0.6 0.2 0.07 13.5 13.9 72.6 48.7

Table 4Comparison between wet and dry tar measurement method used in this research.

Wet method Dry method

Objectiveand outcome

Determine tar removalefficiency in term ofgravimetric tar weight perproducer gas volume (g/m3)

Trace alteration of tarconcentration along withoperating time andinvestigation on the saturatedpoint of tar removal medium

Sampled tarcompound

Heavy tar as gravimetric tarweight

Light tar with 10 standards;benzene, toluene, xylene,stryrene, phenol, indene,naphthalene, phenanthrene,anthracence and pyrene

Preparationtime

Long preparation time No preparation required

Samplingequipment

The chain of ten impingers ofisopropanol has to be preparedin cold baths

The charcoal tube for non-polartar components and silica geltube for polar tar componentswere purchased commercially

Sampling time Long sampling time Short sampling timeAnalysis time Short analysis time Long analysis timeAnalysisequipment

Rotary evaporator is used toseparate isopropanol fromgravimetric tar

GC-FID using carbon disulfideand acetone as solvent is usedto fractionate tar aerosols


was principally for non-polar organic compounds adsorption and thesilica gel tube was principally for polar organic compounds adsorp-tion. The tube train was connected with filters, a gas flow meter, amanual control valve, and a suction pump, respectively. The produc-er gas was sampled at the inlet and exit of the adsorbent bed with theconstant flow rate of 0.5 l/min for 3 min. The sampling was done every6 min correspondingly as shown in Fig. 5 where Ai means gas samplingat the inlet of the adsorbent bed and Bi means gas sampling at the exitof the adsorbent bed. After finishing tar sampling, these sampling tubes

PurgeP

ActivatCarbo

Gas Flow Meter

ManualControl Valve

SuctionPump

Mixture of salt/water/ice bath at

Producer gas

3±1°C

Fig. 4. Tar measurement — wet meth

were kept in a refrigerator to avoid adsorbed tar evaporation. Gas chro-matographyflame ionization detector (GC-FID)was utilized to determinelight tar compounds and their concentrations. Here, carbon disulfide andacetone were used as solvent for charcoal tube and silica gel tube,respectively.

4. Experiment procedure

Initial experimental setup conditions and summarized conditions ofeach experiment run are shown in Tables 5 and 6, respectively. The gascleaning unit was installed between the reformer and the gas samplingline as shown in Fig. 1. The absorption unit was used in Run 1 and Run 2,the adsorption unitwas used in Run 3 and Run4 and the combination ofthe absorption and the adsorption units was used in Run 5.

4.1. Absorption

In this experiment, tar removal performance of two type of scrub-bing medium were investigated; vegetable oil and waste-cooking oil.500 ml of scrubbing mediumwas contained in a three-neck flask witha magnetic stirrer inside. An adjustable digital magnetic stirrer ma-chine controlled the speed of the stirrer. The scrubber flask was con-nected between the reformer and the wet type sampling line. Theinvestigated stirring speeds were 0, 300, 750, 1000 and 1500 rpm.The scrubber was kept at the room temperature.

4.2. Adsorption

The adsorbent was packed in a stainless steel cylinder fixed bedreactor with the dimension of 50 mm I.D.×250 mm height. The reac-tor contained 100 mm height of the adsorbent. The adsorbent bedwas kept at the room temperature.

4.3. Combination of absorption and adsorption

For the experiment combining the absorption and adsorption, ad-sorbent bed was connected downstream of the oil scrubber. The oilscrubber and the adsorbent bed were kept at the room temperature.The wet type sampling line was connected downstream of the ad-sorption reactor. The producer gas passes through the cleaning unitsat the room temperature.

Cooler

ed n

Cotton Filter II

Cotton Filter I

Isopropanol bath at -22±1°C

od for heavy tar measurement.

PurgeP

Activated Carbon

Cotton Filter II

Cotton Filter I

Gas Flow Meter

ManualControl Valve

SuctionPump

Charcoal Tube Silica Gel Tube

20 min 96 min

B1 B2 B3 B4 B5 B6 B7 B8 B9 B10 B11 B12

A1 A2 A3 A4 A56 min

Start Feeder

Experiment Period

Fig. 5. Tar measurement — dry method for light tar measurement.

Table 6Experimental conditions of each experiment run.

Parameters Run 1 Run 2 Run 3 Run 4 Run 5

Absorbentscrubber

Vegetable oil Waste-cooking oil

– – Vegetableoil

Scrubber volume(ml)

500 500 – – 500

Scrubber mixingspeed (rpm)

0,300,750,1000,1500

0,300,750,1000,1500

– – 0, 1000

Adsorbent bed – – Ricehusk

Ricehuskchar

Rice huskchar


5. Result and discussion

5.1. Comparison of scrubbing performance of vegetable oil and waste-cooking oil for tar removal

Due to non-polar characteristic of the oil solution, tar compoundsare able to dissolve in the oil and that makes the oil scrubber capableof tar removal [21]. This work compared the gravimetric tar removal ef-ficiency of the vegetable oil and waste-cooking oil scrubbers under thesame operating conditions. Fig. 6 presents the vegetable oil and waste-cooking oil scrubber performances for the gravimetric tar removal. Itcan be seen that 93.7 g/m3 of the gravimetric tar contained in the pro-ducer gas was reduced to 34.1 g/m3 which corresponds to 63.6% gravi-metric tar removal efficiency and 40.8 g/m3which corresponds to 56.4%gravimetric tar removal efficiency by using the vegetable oil andwaste-cooking oil scrubbers, respectively. Therefore, the vegetable oil scrubbershowed better performance for tar removal compared to the wastecooking oil scrubber.

After experiencing the oxidative, hydrolytic and thermolytic reac-tions when fried or deep-fried at approximately 150 to 450 °C, the mo-lecular structure of used cooking oil will be changed from fresh cookingoil [38–40]. It will become non-uniform and contaminated with con-taminants such as oxygen, hydrogen, carbon and nitrogen comingfrom food, which are not able to be removed by pre-treatment process-es. These contaminants in the waste-cooking oil contributed to the de-crease of the absorption efficiency of tar by blocking the absorbentsurface area of waste-cooking oil molecules. Therefore, solubility of tarin the waste-cooking oil decreased. Moreover, from the comparison ofthe characteristics of the both oils shown in Table 2, the vegetable oilhas lower viscosity than the waste-cooking oil. For that reason, the dif-fusivity of tar in the vegetable oil is higher. Therefore, the vegetable oilhas higher tar removal efficiency.

5.2. Turbulent mixing effects of vegetable oil and waste-cooking oilscrubbers on tar removal

Turbulent mixing effect of the oil scrubbers was investigated. Themixing speed of 300, 750, 1000 and 1500 rpm were examined for

Table 5Initial experimental setup conditions.

Parameters

Pyrolyzer (°C) 800Reformer (°C) 800Carrier gas N2

Carrier gas rate (l/min) 1.5Feedstock Rice huskFeedstock size (mm) 0.125–0.500

both the vegetable oil and the waste-cooking oil. The results areshown in Fig. 7.

5.2.1. Vegetable oil scrubberFor the vegetable oil scrubber, the tar removal efficiency increased

with the increase of the mixing speed then decreased when the mixingspeed was higher than 1000 rpm. As a result, the best condition wasgiven at the 1000 rpm mixing speed, where tar could be reduced to9.6 g/m3, corresponded to 89.8% gravimetric tar removal efficiency.These results can be explained by three main factors; bubble size, con-tacting time and theVanderWaal's force betweenoil and tarmolecules.As the mixing speed increased, oil bubbles were getting smaller,resulted in the increase of the oil surface area and that gave more areaon oil molecule surface for tar to be absorbed. On the other hand,when the stirring speed was too high, the contacting time between tarand scrubbing medium decreased and the Van Der Waal's forcebetween tar and oil molecule bonding got weaker. Therefore, becauseof a high mixing speed, some absorbed tar, which coagulated on oilmolecules, de-absorbed and went out with the flowing out gas. Subse-quently, less tar could be absorbed by the absorbent.

5.2.2. Waste-cooking oil scrubberFor the waste-cooking oil scrubber, the tendency of tar removal per-

formance at eachmixing speedwas similar to vegetable oil. The optimum

Adsorbent meshsize (mm)

– – 0.5–1.0

0.5–1.0 0.5–1.0

Adsorbent height(mm)

– – 100 100 100

MediumTemperature (°C)

27–28 27–28 27–28

27–28 27–28

Tar measurement Heavy tar Heavy tar Lighttar

Light tar Heavy tar

Tar measurementmethod

Wet Wet Dry Dry Wet

Sampling time(min)

48 48 96 96 48

0

20

40

60

80

100

No gas Vegetable Waste-cookingcleaning Oil Oil

Gra

vim

etric

Tar

[g/m

3 ]

0 20 40 60 80 100

% Gravimetric Tar Removal

VegetableOil

Waste-cookingOil

Fig. 6. Comparison of the vegetable oil and waste-cooking oil scrubber performances for tar removal.


condition was at 750 rpm mixing speed where tar could be reduced to17.4 g/m3, which corresponded to 81.4% gravimetric tar removal efficien-cy. The results can be explained by four factors; bubble size, contactingtime, the Van der Waal's force between oil and tar molecules, and con-taminants in waste-cooking oil. As the mixing speed increased, oil bub-bles were getting smaller while bonding force between contaminantsand oil molecules got weaker and made contaminants de-absorbed andflew away with the flowing out gas. Therefore, surface area on oil mole-cules increased and that gave more area on oil molecules for tar to beabsorbed. On the contrary, when the stirring speed was too high, the tarremoval efficiency decreased because of the decrease of the contactingtime andweakening of the Van derWaal's force between oil and tar mol-ecules as explained above in the vegetable oil case.

5.2.3. Comparison of tar removal performance at each mixing speed be-tween vegetable oil and waste-cooking oil scrubbers

It was indicated that, in all conditions, vegetable oil showed bettertar removal performance and the optimum mixing speed was1000 rpm for the vegetable oil scrubber and 750 rpm for the waste-cooking oil. This can be explained that, contaminants in waste-cooking oil, such as nitrogen and sulfur [40], weaken the Van derWaal's force between oil molecules and absorbed tar. Therefore,with a too high mixing speed, not only contaminants but alsoabsorbed tar would de-absorb and flew out with the flowing out gas.

5.3. Comparison of tar adsorption performance of rice husk and rice huskchar

The tar adsorption performances of rice husk char and rice huskadsorbents were observed by breakthrough curves for each tar com-ponent as shown in Figs. 8 and 9, respectively. The tar amount inthe producer gas at the inlet and exit positions of the adsorbent bedwas measured. The examined tar components consist of light aromat-ic hydrocarbon tars, which are benzene, toluene, xylene, styrene,phenol and indene, and light poly aromatic hydrocarbon (PAH) tars,which are naphthalene, phenanthrene, anthracene and pyrene.

0

20

40

60

80

100

Vegetable OilWaste-cooking Oil

No Scrubber

0 300 750 1000 1500mixing speed [rpm]

Gra

vim

etric

Tar

[g/m

3 ]

Fig. 7. Turbulent mixing effect of the vegetable oil a

5.3.1. Rice husk char adsorbentThe adsorption performance of rice husk char for light aromatic

hydrocarbon tars and light PAH tars was visibly examined by break-through curves as shown in Fig. 8. Some of light aromatic hydrocar-bon tars; xylene and styrene were well adsorbed along with thesampling time but for a short period due to saturated adsorbentarea on rice husk char surface area. Consequently the poor adsorptionperformance afterward and higher amount of tars was found due tothe de-adsorbed tars surged out with the exit gas. The same tendencyresults were obtained for indene and naphthalene but for the longerbreakthrough time, compared with xylene and styrene, which wasapproximately 1 hour. Phenanthrene, anthracene and pyrene weremoderately adsorbed by rice husk char throughout the samplingtime. Phenol was found to be greatly removed after adsorption. Thiscould be explained by the water-soluble characteristic of phenol. Phe-nol was dissolved in water generated from biomass and was filteredin the adsorbent bed.

The good adsorption performance of rice husk char can be clearlyexplained by the high BET surface area value shown in Table 3 andSEM micrograph of its porous texture shown in Fig. 3.

As mentioned that rice husk char is proposed for light tar removal.Therefore, the adsorption performance may not rapidly decrease andthe adsorption will reach saturation point in a long period of opera-tion. Nevertheless, due to the adsorption bed was at the ambient tem-perature, slight tar was found condensed and filtered in the adsorbentbed.

In large scale operation, It is highly suggested to investigate the sat-uration point of rice husk char adsorbent under each operating condi-tion individually before designing the gas cleaning system. Once thechar adsorption capacity is saturated after tar adsorption, it should betreated properly. In general, there are several ways to treat or regener-ate the used adsorbent for an economically feasible application such asthermal, extractive and chemical regeneration process [41,42]. On thecontrary, the main aim of this research is to utilize the waste char thatinitially integrated in the gasification process. The weight distributionof char production from gasification process at 800 C is approximately

40

50

60

70

80

90

100

Vegetable Oil

Waste-cooking Oil

0 300 750 1000 1500mixing speed [rpm]

% G

ravi

met

ric T

ar R

emov

al

nd waste-cooking oil scrubbers on tar removal.

0

5

10

15

0 24 48 72 96

Benzene

0

0.5

1

1.5

2

0 24 48 72 96

Toluene

0

0.025

0.05

0.075

0.1

0 24 48 72 96

Xylene

0

0.25

0.5

0.75

1

0 24 48 72 96

Styrene

0

0.025

0.05

0.075

0.1

0 24 48 72 96

Phenol

0

0.0125

0.025

0.0375

0.05

0 24 48 72 96

Phenanthrene

0

0.2

0.4

0.6

0 24 48 72 96

Indene

0

0.2

0.4

0.6

0.8

0 24 48 72 96

Naphthalene

0

0.005

0.01

0.015

0.02

0 24 48 72 96

Anthracene

0

0.001

0.002

0.003

0.004

0 24 48 72 96

Pyrene

Experiment Period [min]C

once

ntra

tion

[g/m

3 ]

Con

cent

ratio

n [g

/m3 ]

Con

cent

ratio

n [g

/m3 ]

Con

cent

ratio

n [g

/m3 ]

Con

cent

ratio

n [g

/m3 ]

Con

cent

ratio

n [g

/m3 ]

Con

cent

ratio

n [g

/m3 ]

Con

cent

ratio

n [g

/m3 ]

Con

cent

ratio

n [g

/m3 ]

Con

cent

ratio

n [g

/m3 ]

Experiment Period [min]

Experiment Period [min] Experiment Period [min]

Experiment Period [min]Experiment Period [min]



Fig. 8. Tar removal performance of rice husk char; ♦ indicates gas sampling at the inlet of the adsorbent bed (Ai) and ■ indicates gas sampling at the exit of the adsorbent bed (Bi).


at 10% while liquid and gas production is at 5% and 85%, respectively[43]. This 10% char production will be utilized as tar removal adsorbentin the gas cleaning process. The regenerate process of tar adsorbed charto remove volatile matter as tar may effectively redevelop the porestructure of char and restore the char capacity, butmay not worth com-paring the quality of waste rice husk char with the costly regenerationprocess. The most suitable to handle used char adsorbent is to utilize

as fuel in the gasifier at the initial heat-up process by mixing with for-mer heating material such as silica sand.

5.3.2. Rice husk adsorbentFrom the breakthrough curve of tar removal performance of rice

husk shown in Fig. 9, it was obviously seen that rice husk had less ad-sorption capacity for light aromatic hydrocarbon tars and light PAH






0

5

10

15

0 24 48 72 96

Benzene

0

0.5

1

1.5

2

0 24 48 72 96

Toluene

0

0.2

0.4

0.6

0.8

0 24 48 72 96

Styrene

0

0.05

0.1

0.15

0.2

0 24 48 72 96

Phenol

0

0.2

0.4

0.6

0 24 48 72 96

Indene

0

0.2

0.4

0.6

0.8

1

0 24 48 72 96

Naphthalene0

0.015

0.03

0.045

0.06

0 24 48 72 96

Phenanthrene

0

0.01

0.02

0.03

0.04

0 24 48 72 96

Anthracene

0

0.0025

0.005

0.0075

0.01

0 24 48 72 96

Pyrene

0

0.025

0.05

0.075

0.1

0 24 48 72 96

Xylene

Con

cent

ratio

n [g

/m3 ]

Con

cent

ratio

n [g

/m3 ]

Con

cent

ratio

n [g

/m3 ]

Con

cent

ratio

n [g

/m3 ]

Con

cent

ratio

n [g

/m3 ]

Con

cent

ratio

n [g

/m3 ]

Con

cent

ratio

n [g

/m3 ]

Con

cent

ratio

n [g

/m3 ]

Con

cent

ratio

n [g

/m3 ]

Con

cent

ratio

n [g

/m3 ]

Fig. 9. Tar removal performance of rice husk; ♦ indicates gas sampling at the inlet of the adsorbent bed (Ai) and ■ indicates gas sampling at the exit of the adsorbent bed (Bi).


tars. Light aromatic hydrocarbon tars were slightly decreased alongthe sampling time. Light PAH tars removal has found to be betterthan light aromatic hydrocarbon tars removal with more amountswas decreased along the sampling time. Therefore, it is believedthat light PAH tars which are in the class of condensable tar werenot removed by rice husk adsorption but by condensation at lowertemperature of the adsorbent bed. This observation is well agreedwith the consideration of BET surface area and SEM micrograph that

showed low surface area of rice husk and the absence of porosity onthe rice husk surface for the adsorption process.

5.4. Combination of vegetable oil scrubber with rice husk char adsorptionbed

From the results of absorption and adsorption techniques, it wasfound that the optimum absorption condition was vegetable scrubber

40

50

60

70

80

90

100

0

20

40

60

80

100

No cleaning unit

% G

ravi

met

ric T

ar R

emov

al

No Veg oil Veg oil Veg oilmixing + char 1000 rpm 1000 rpm

+ char

No Veg oil Veg oil Veg oil mixing + char 1000 rpm 1000 rpm

+ char

Gra

vim

etric

Tar

[g/m

3 ]

Fig. 10. Tar removal performance of the combination of vegetable oil scrubber with rice husk char adsorption bed at each case.


with 1000 rpm mixing speed and the optimum adsorbent was ricehusk char. Therefore, the experiment of combination of the vegetableoil scrubber with no mixing and 1000 rpm mixing speed combinedrice husk char adsorption was set up and achieved 89.8% and 95.4%gravimetric tar removal efficiency, respectively, as shown in Fig. 10.The tar removal efficiency was almost the same in the cases of vege-table oil scrubbing without stirring combined with adsorption andvegetable oil scrubbing with 1000 rpm mixing speed. These resultsshow an interesting fact that the combination system showed thehighest tar removal efficiency with the economically feasibilityusing low-cost char produced in the gasification process as an effec-tive carbon adsorbent, which was a bonus advantage from gasifica-tion system itself.

6. Conclusions

Two kinds of tar were concerned in this research; heavy and lighttar. This research is aimed for developing the most effective tar re-moval system, which is able to remove both kinds of tar. Since the na-ture of heavy and light tar is different, two individual tar removaltechniques were investigated. As oil is non-polar substance, heavytar is proposed to be absorbed by oil, while light tar would beadsorbed by rice husk char adsorbent due to its high porosity proper-ty. Tar absorption performances of vegetable oil scrubbing and waste-cooking oil scrubbing were compared and tar adsorption perfor-mances of rice husk and rice husk char were compared. A laboratoryscale pyrolyzer was used to produce tar rich producer gas from ricehusk feedstock. For absorption, vegetable oil and waste-cooking oilwere found to be effective absorbent for gravimetric tar removal. Itwas found that the combination of these two techniques created pos-itive synergetic effect that heavy tar, which is mostly oily liquid orsolid at ambient temperature, will be firstly absorbed and removedby vegetable oil scrubber then the remain light tar in the producergas, which is mostly in gas phase, will be adsorbed and removed byrice husk char and release gratified tar-free producer gas to the down-stream application. To enhance the performance of heavy tar remov-al, turbulent mixing oil scrubber was investigated and found to bevery effective in lab scale. In large scale operation, this combinationof tar removal technique is suggested to set up in series of circulatedoil scrubber connected with rice husk char filter tower. Turbulentmixing showed significant improvement of the absorption efficiencyby reducing the bubble size and increasing the absorbent surfacearea. In this research, the vegetable oil showed better tar absorptionperformance than the waste-cooking oil, and 1000 rpm was the opti-mum mixing speed for the vegetable oil and 750 rpm was the opti-mum mixing speed for the waste-cooking oil. For adsorption, ricehusk char showed good performance in the term of light tar adsorp-tion due to its porous property. SEM and BET studies confirmed thetar adsorption ability of rice husk char. The combination of the vege-table oil scrubber with the rice husk char adsorbent bed could removeas high as 95.4% of the gravimetric tar. The remarkable findings were

the turbulent mixing of oil scrubber which may be circulated oilscrubber for heavy tar removal in the large scale operation, the capa-bility of waste char for light tar adsorption and the combination ofthese two promising techniques for the optimum tar removal systemunder ambient temperature.

References

[1] W. Zhao, Z.Q. Li, D.W. Wang, Q.Y. Zhu, R. Sun, B.H. Meng, G.B. Zhao, Combustioncharacteristics of different parts of corn straw and NO formation in a fixed bed,Bioresource Technology 99 (2008) 2956–2963.

[2] P. McKendry, Energy production from biomass (Part1): overview of biomass, Bio-resource Technology 83 (2002) 37–46.

[3] O. Senneca, Kinetic of pyrolysis, combustion and gasification of three biomassfuels, Fuel Process Technology 88 (2007) 87–97.

[4] Juan Daniel Martinez, Khamid Mahkamov, Rubenildo V. Andrade, Electo E. SilvaLora, Syngas production in downdraft biomass gasifiers and its applicationusing internal combustion engines, Renewable Energy 38 (2012) 1–9.

[5] A.V. Bridgwater, D. Meier, D. Radlein, An overview of fast pyrolysis of biomass, Or-ganic Geochemistry 30 (1999) 1479–1493.

[6] A.V. Bridgwater, Renewable fuels and chemicals by thermal processing of bio-mass, Chemical Engineering Journal 91 (2003) 87–102.

[7] Digman Brett, Joo Hyun Soo, Kim Dong-Shik, Recent progress in gasification/ py-rolysis technologies for biomass conversion to energy, American Institute ofChemical Engineers, Environmental Progress & Sustainable Energy 28 (1)(2009) 47–51.

[8] M.A. Caballero, J. Corella, M.P. Aznar, J. Gil, Biomass gasification with air in fluid-ized bed. Hot gas cleanup with selected commercial and full-size nickelbased cat-alysts, Industrial and Engineering Chemistry Research 39 (2000) 1143–1154.

[9] R. Warnecke, Gasification of biomass: comparison of fixed bed and fluidized bedgasifier, Biomass and Bioenergy 18 (2000) 489–497.

[10] Sang jun Yoon, Yung-II Son, Yong-Ku Kim, Jae-Goo Lee, Gasification and powergeneration characteristics of rice husk and rice husk pellet using a downdraftfixed-bed gasifier, Renewable Energy 42 (2012) 163–167.

[11] A.K. Jain, J.R. Goss, Determination of reactor scaling factors for throatless rice huskgasifier, Biomass and Bioenergy 18 (2000) 249–256.

[12] L. Devi, K.J. Ptasinski, F.J.J.G. Janssen, S.V.B. van Paasen, P.C.A. Bergman, J.H.A. Kiel,Catalytic decomposition of biomass tars: use of dolomite and untreated olivine,Renewable Energy 30 (2005) 565–587.

[13] K. Maniatis, A.A.C.M. Beenackers, Introduction: tar protocols., IEA gasificationtasks, Biomass and Bioenergy 18 (2000) 1–4.

[14] P.C.A. Bergman, S.V.B. van Paasen, H. Boerrigter, The novel “OLGA” technology forcomplete tar removal from biomass producer gas, Pyrolysis and gasification ofbiomass and waste, expert meeting, 2002 France: Strasbourg.

[15] H.W.J. Könemann, S.V.B. van Paasen, OLGA tar removal technology; 4 MW com-mercial demonstration, 15th European Biomass Conference and Exhibition, Ber-lin, Germany, 2007.

[16] C. Li, K. Suzuki, Tar property, analysis, reforming mechanism and model for bio-mass gasification — an overview, Renewable and Sustainable Energy Reviews13 (2009) 594–604.

[17] L. Devi, K.J. Ptasinski, F.J.J.G. Janssen, A review of the primary measures for tarelimination in biomass gasification processes, Biomass and Bioenergy 24 (2003)125–140.

[18] S. Anis, Z.A. Zainal, Tar reduction in biomass producer gas via mechanical, catalyt-ic and thermal methods: A review, Renewable and Sustainable Energy Reviews15 (2011) 2355–2377.

[19] M. Dogru, A. Midilli, C.R. Howarth, Gasification of sewage sludge using a throateddowndraft gasifier and uncertainty analysis, Fuel Processing Technology 75(2002) 55–82.

[20] S. Koppatz, C. Pfeifer, R. Rauch, H. Hofbauer, T. Marquard-Moellenstedt, M. Specht,H2 rich product gas by steam gasification of biomass with in situ CO2, Fuel Pro-cessing Technology 90 (2009) 914–921.

[21] T. Phuphuakrat, T. Namioka, K. Yoshikawa, Absorptive removal of biomass tarusing water and oily materials, Bioresource Technology 102 (2011) 543–549.


[22] R.W.R. Zwart, Gas cleaning downstream biomass gasification Status Report ECN-E–08-078, , 2009.

[23] S. Majumdar, D. Bhaumik, K.K. Sirkar, G. Simes, A pilot-scale demonstration of amembrane-based absorption- stripping process for removal and recovery of vol-atile organic compounds, Environmental Progress 20 (2001) 27–35.

[24] B. Ozturk, D. Yilmaz, Absorptive removal of volatile organic compounds from fluegas streams, Process Safety and Environment Protection 84 (2006) 391–398.

[25] C. Pfeifer, R. Rauch, H. Hofbauer, In-bed catalytic tar reduction in a dual fluidizedbed biomass steam gasifier, Industrial and Engineering Chemistry Research(2004) 1634–1640.

[26] Arjun B. Chhetri, K. Chris Watts, M. Rafiqul Islam, Waste cooking oil as an alterna-tive feedstock for biodiesel production, Energies 1 (2008) 3–18.

[27] T. Phuphuakrat, T. Namioka, K. Yoshikawa, Tar removal from biomass pyrolysisgas in two-step function of decomposition and adsorption, Applied Energy 7(2010) 2203–2211.

[28] X. Hu, T. Hanaoka, K. Sakanishi, T. Shinagawa, S. Matsui, M. Tada, Removal of tarmodel compounds produced from biomass gasification using activated carbons,Journal of the Japan Institute of Energy 96 (2007) 707–711.

[29] M.A. Lillo-Ródenas, A.J. Fletcher, K.M. Thomas, D. Cazorla-Amorós, A. Linares-Solano, Competitive adsorption of a benzene–toluene mixture on activated car-bons at low concentration, Carbon 44 (2006) 1455–1463.

[30] A.M. Mastral, T. Garcia, M.S. Callen, M.V. Navarro, J. Galban, Assessment of phen-anthrene removal from hot gas by porous carbons, Energy & Fuels 15 (2001) 1–7.

[31] A.M. Mastral, T. Garcia, R. Murillo, M.S. Callen, J.M. Lopez, M.V. Navarro, Measurementsof polycyclic aromatic hydrocarbon adsorption on activated carbons at very low con-centrations, Industrial and Engineering Chemistry Research 42 (2002) 155–161.

[32] Z. Abu El-Rub, E.A. Bramer, G. Brem, Experiment comparison of biomass chars withother catalysts for tar reduction, Fuel 87 (2008) 2243–2252.

[33] DuoWang, Wenqiao Yuan, Wei Ji, Char and char-supported nickel catalysts for sec-ondary syngas cleanup and conditioning, Applied Energy 88 (2011) 1656–1663.

[34] Yue Chen, Yanchao Zhu, Zichen Wang, Ying Li, Lili Wang, Lili Ding, Xiaoyan Gao,Yuejia Ma, Yupeng Guo, Application studies of activated carbon derived from

rice husks produced by chemical–thermal process — A review, Advances in Col-loid and Interface Science 163 (2011) 39–52.

[35] P. Gilbert, C. Ryu, V. Sharifi, J. Swithenbank, Tar reduction in pyrolysis vapours frombiomass over a hot char bed, Bioresource Technology 100 (2009) 6045–6051.

[36] A. van der Drift, M.C. Carbo, C.M. van der Meijden, The TREC-module: integrationof tar reduction and high-temperature filtration, Proceedings of the 14th Europe-an Biomass Conference and Exhibition, Paris, France, 17–21 October 2005, 2005.

[37] ECN-C-02-090, Guideline for sampling and analysis of tar and particles in biomassproducer gas, , November 2002.

[38] C.W. Yu, S. Bari, A. Ameen, A comparison of combustion characteristics of wastecooking oil with diesel as fuel in a direct injection diesel engine, Proceedings of theInstitute of Mechanical Engineering, Part D: Journal of Automobile Engineering216 (2002) 237–243.

[39] J.M. Nzikou, L. Matos, J.E. Moussounga, C.B. Ndangui, N.P. Pambou-Tobi, E.M.Bandzouzi, A. Kimbonguila, M. Linder, S. Desobry, Study of oxidative and thermalstability of vegetable oils during frying, Research Journal of Applied Science 4 (2)(2009) 94–100.

[40] Stella Bezergianni, Spyros Voutetakis, Aggeliki Kalogianni, Catalytic hydrocrack-ing of fresh and used cooking oil, Industrial and Engineering Chemistry Research48 (2009) 8402–8406.

[41] J. Carratala-Abril, M.A. Lillo-Rodenas, A. Linares-Solano, D. Cazorla-Amoros, Regenera-tion of activated carbons saturated with benzene or toluene using an oxygen-containing atmosphere, Chemical Engineering Science 65 (2010) 2190–2198.

[42] G. San Miguel, S.D. Lambert, N.J.D. Graham, The regeneration of field-spent gran-ular activated carbon, Water Research 35 (2001) 2740–2748.

[43] J.A. Libra, K.S. Ro, C. Kammann, A. Funke, N.D. Berge, Y. Neubauer, M.-M. Titirici,C. Fuhner, O. Bens, J. Kern, K.-H. Emmerich, Hydrothermal carbonization of bio-mass residuals; a comparative review of the chemistry, processes and applica-tions of wet and dry pyrolysis, Biofuels 2 (1) (2011) 89–124.

pa ethan om 2012

Documents