akiŞkan yatakli reaktÖrde kÖmÜr pİrolİzİ

7/30/2019 AKIKAN YATAKLI REAKTRDE KMR PROLZ

1/10

Journal of Analytical and Applied Pyrolysis 91 (2011) 241250

Contents lists available at ScienceDirect

Journal of Analytical and Applied Pyrolysis

j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / j a a p

Coal pyrolysis in a fluidized bed reactor simulating the process conditions of coal

topping in CFB boiler

Xiaofang Zhang, Li Dong, Juwei Zhang, Yajun Tian, Guangwen Xu

State Key Laboratory of Multi-Phase Complex System, Institute of Process Engineering, Chinese Academy of Sciences, ZhongGuanCun Beiertiao 1,

Haidian District, Beijing 100080, China

a r t i c l e i n f o

Article history:

Received 14 September 2010

Accepted 18 February 2011

Available online 26 February 2011

Keywords:

Coal topping

CFB boiler

Fluidized bed

Reaction atmosphere

Pyrolysis oil

TG-FTIR

a b s t r a c t

Simulating the conditions of pyrolytic topping in a fluidized bed reactor integrated into a CFB boiler, the

study was devoted to the reaction fundamentals of coal pyrolysis in terms of the production character-

istics of pyrolysis oil in fluidized bed reactors, including pyrolysis oil yield, required reaction time and

the chemical species presented in the pyrolysisoil. The results demonstrated that themaximal pyrolysis

oil yield occurred on conditions of 873 K, with a reaction time of 3 min and in a reaction atmosphere gas

simulating the composition of pyrolysis gas. Adding H2 and CO2 into the reaction atmosphere decreased

thepyrolysisoil yield, while theoil yield increasedwith increasing theCO and CH4 contents in theatmo-

sphere. TG-FTIR analysis was conducted to reveal the effects of reaction atmosphere on the chemical

species present in the pyrolysisoil. The results clarified that the pyrolysisoil yield reached its maximum

when the simulated pyrolysis gas was the reaction atmosphere, but there were slightly fewer volatile

matters in the pyrolysis oil thanthe oil generated in the N2 atmosphere. All of these results are expected

notonly to revealthe composition characteristicsof thepyrolysis oilfrom differentconditions of thecoal

topping process but also to optimize the pyrolysis conditions in terms of maximizing the light pyrolysis

oil yield and quality.

2011 Elsevier B.V. All rights reserved.

1. Introduction

Coal topping (pyrolytic topping) process was proposed by Yao

and Kwauk[1,2] to achieve high-value utilization of coal consumed

in CFB boilers. As shown schematically in Fig. 1, light liquid product

in this process is produced by flash pyrolysis of coal in a pyrolyzer

integrated into a circulating fluidized bed boiler which burns the

pyrolysis-generated char to generate heat and electricity. Previous

studies showedthat the pyrolysisis necessary to proceed withrapid

heating andin turn thequickseparation(from solids) andquench of

the gaseous product in order to minimize the secondary reactions

like cracking and polymerization [2,3].

A few, although limited studies have been done regarding

pyrolytic topping. Wang et al. [4] found that downer is a suitable

reactor for implementing thepyrolysis of coal through mixing with

thehot ashparticles duringtheirfallin thereactorvia gravity.Those

authors further investigated the influences of pyrolysis tempera-

ture and particle size on the topping performance [5,6]. Because

the coal particle residence time is short (only a few seconds), the

downer reactor only adapts to coal in micrometers. Moving bed is

Corresponding author. Tel.: +86 10 62550075; fax: +86 10 62550075.

E-mail address: [email protected](G. Xu).

another kind of reactor used to implement the pyrolytic topping,

and Bi [7] employed this reactor to integratethe coal pyrolysiswith

a riser combustor. Comparingto thedowner reactor, a special effort

is needed to scale up the reactor to achieve high-efficiency mixing

with coal particles for industrial application. Fluidized bed allows

easier scale-up and also adapts wide-size particles (


2/10

242 X. Zhang et al. / Journal of Analytical and Applied Pyrolysis 91 (2011) 241250

Fig. 1. Principle of the coal pyrolytic topping process in CFB boiler.

Fig. 2. Schematic diagram of the adopted experimental apparatus.

the sand bed in the reactor was adjusted according to the experi-

mental needs. A three-zone electric furnace heated the reactor and

the heating conditions for each zone could be independently con-

trolled. A thermocouple was immersed in the quartz sand bed to

monitor and measure the temperature of the fluidized bed. The

reaction atmosphere (N2 or mixture of H2, CO, CO2 and CH4) was

formed by mixing gases from different cylinders and the flow ratewas kept at 1.15 m/min (2 times of minimum fluidization velocity)

to achieve the full fluidization of the quartz sand particles. The par-

ticle sizes of the tested lignite were 46 mm, and Table 1 shows the

major properties of the tested coal.

Table 1

Proximate and ultimate analyses of the tested coal.

Pr oximate an alys is [db- wt. %] Ultimate analysis [db- wt.% ]

Water (arrival base): 1.7 C: 65.6

Volatile: 31.2 H: 4.1

Ash: 18.2 S: 0.6

Fixed C: 48.9 N: 1.1

LHV [MJ kg1] 24.96 O: 9.7

Table 2

Typical FTIR absorption peaks and their implicated functional groups [8].

Absorbance [cm1] Function group Characteristic chemicals

13551395, 14301470 CH3 Methyl

14051465 CH2 Methylene

1500 C C Single ring aromahydrocarbon

1740 C O Acid ketone

28003100 CH4 Methane

20002250 CO Carbon monoxide

22502400 CO2 Carbon dioxide2920 CH3, CH2, CH Aliphatic

3500 OH(a) Phenolic hydroxyl

3650 OH(b) Alcoholic hydroxyl

The reactor was first heated to 673 K before the fluidizing gas

(N2) was introduced into the reactor. The bed was heated to the

desired temperature in N2 atmosphere and then the fluidizing gas

was switched to the required reaction atmosphere. When the bed

temperature in the new atmosphere reached the specified steady

value, 10 g of lignite were added into the reactor from the bed top

through a valve-hopper. The generated pyrolysis gas was cooled

immediately in a water cooler and then washed with water and

acetone in succession cooled via an ice-water bath. The volume of

the gas was measured by a wet gas meter after the washing bath.After passing through a filter and drier further, the cleaned gas was

sampled at the end of the gas line at an interval of about 20s. At

the end of gas sampling, the gas from the reactor was switched to a

bypass line to vent without passing through the above-mentioned

gas cooling and washing vessels. Therefore, the time to sample the

gas also represented the reaction time measured for the pyrolysis,

which was usually a few minutes after the coal feeding. The sam-

pled gaswas analysed using a micro GC(Agilent3000) to determine

its composition.

The liquid collected from washing the cooler and pyrolysis gas

was treated to recover pyrolysis oil (tar) through filtration and

remove both acetone and water. The collected liquid was first

treated in an atmospheric rotary evaporator at 318 K and then

in a vacuum oven at 318K. The quality of the pyrolysis oil wasdetermined by a thermal gravity analyzer integrated with a Fourier

transform infrared spectrometer (FTIR).

2.2. Analysis approach and mass balance

A micro GC was used to measure the molar concentrations of

H2, O2, N2,CO,CO2 and hydrocarbons up to C3 in the gaseous prod-

uct. In order to gain composition information of pyrolysis oil, a

thermo gravimetric analyzer (Netzsch STA 449C) coupled with a

FTIR (i.e., TG-FTIR) was employed to analyse the recovered pyroly-

sis oil. In the TG-FTIR analysis, the pyrolysis oil sample was heated

from 303 K to 1173K at 30K/min in TG. Nitrogen at 80mL/min was

adopted to carry the evolved volatile gas from TG to the gas cell of

FTIR heated to 573 K. The FTIR analysis was conducted at TG tem-peraturesbetween 373K and1273K. Theresulting intensity of FTIR

spectrum was mass-normalized to eliminate the influence of sam-

ple mass. Table 2 lists the characteristic peaks in FTIR spectra and

their implicated functional groups. The yield of pyrolysis oil (Yoil,

wt.%) and the production rate of pyrolysis gas (Ygas, L/g) with a dry

ash-free basis were calculated with

Yoil =moil

mcoal (100Mad Aad) 100% (1)

and

Ygas =

22.4

i

t0FmtCidt

mcoal

(100MadA

ad) 100% (2)


3/10

X. Zhang et al. / Journal of Analytical and Applied Pyrolysis 91 (2011) 241250 243

Fig. 3. Carbon balance examination for the experimental system via gasification

(temperature: 1133K, atmosphere: air+ H2O).

respectively, where Mad and Aad are the contents of water and ash

in coal (wt.%), moil and mcoal are the masses of the produced pyrol-

ysis oil and coal fed into the reactor (kg), Ci is the concentration of

gas species i in pyrolysis gas (vol%), tis the time that the pyrolysis

reaction lasts for or the time for collecting pyrolysis oil (s), and Fmtis the mole flow rate of gas at the time t(mol/s).

Fig. 3 shows the time-series concentrations of the C-containing

gas components (CO, CO2 and CH4) and the correspondingly cal-

culated accumulative C conversion obtained in a gasification test

that was particularly performed to examine the mass balance in

the adopted experimental system. The gasification was tested at

Fig. 4. Time seriesof theconcentrations formajor gas components at thestatic bed

heights of 85mm and 350 mm (temperature: 923K, atmosphere: N2).

Fig. 5. Effectof reaction time on yieldsof pyrolysisoil and gas(temperature:923K,

atmosphere: N2).

1133K in a gas mixture of air (80 vol%) and steam (20vol%). One

can see that the release of CO and CH4 was only at a short ini-

tial stage, denoting actually the period with coal pyrolysis. Then,the char gasification and in turn the combustion of the formed

combustible gas (CO) led to the long-time release of CO2. The accu-

mulative C conversion, defined from the molar ratio of the released

gaseous C over the fed C in the coal, demonstrates that up to 93.0%

C was present in the gas product within the tested 3500 s. The ash

collected at the reactor exit by a filter was found to contain about

2.8% of the fed C. Consequently, the C balance in this test reached

about 96%, showing a good reliability of the testing system. The

result implies that the measurement error would be in 4% in this

article, and this error also represented the repeatability of the tests

reported herein.

3. Results and discussion

In terms of maximizing the pyrolysis oil yield and meanwhile

understanding the oil composition features, the tests were per-

formed to identify the suitable reaction time firstly and then to

clarify the influences of reaction temperature and atmosphere on

the oil yield and composition features.

3.1. Necessary reaction time determination

By presetting the pyrolysis temperature at 923 K, the yields of

gas and oil in pyrolysis were measured in N2 to determine the nec-

essarily required reaction time or the time from feeding coal into

the reactor. Figs. 4 and 5 show the production characteristics of

pyrolysis gas and oil varying with the reaction time. The tests were

performed at two different static bed heights, 85mm and 350 mm,but the analysis in this section will be based on the data from the

350-mm bed test exclusively.

The pyrolysis gas concentrations (excluding N2) varying with

the reaction time in Fig. 4 clarify that the concentrations for all the

gaseous components had their peaks at certain time. Before the

peaks the gradually increased concentrations with prolonging the

reaction time show essentially the quickly deepened degree of the

pyrolysis reaction,while thesuccessive decrease in thegas concen-

trations denotes actually the approach to the end of the pyrolysis

reaction.From the gasconcentration profiles in Fig. 4 one can judge

that the pyrolysis at 923K would finish in about 200s to allow the

major part of the coal volatiles to be released.

Fig. 4 clarifies also that the pyrolysis gas wasrich in CH4 andhad

its lowest concentration for CO2. In-between, both CO and H2 had


4/10


Fig. 6. TG-FTIR spectra of pyrolysis oil produced with different reaction times.

the equivalent concentrations, while all hydrocarbons other thanCH4 manifested a total concentration higher than that of CO 2 but

lower thanthat ofCO and H2. The time to appear the concentration

peakwas differentfor differentgas components. It was equivalently

earliest for CH4 and CO, and in succession it was hydrocarbons

(excluding CH4) and H2. The peak for CO2 was shown up nearly

at the reaction end. These different gas concentration and release

features for different gas species revealed essentially the differ-

ent mechanisms for forming different gas species in coal pyrolysis

[913].

It was reported that in pyrolysis the disruption of HH bonds

occurs first to produce radical H, and the polymerization of the free

radicals H generatesthe molecule H2. Hydrocarbonscan be derived

from cracking fatty matters and aliphatic side-chains of aromatic

molecules. The cracking of aliphatic and aromatic compounds con-

taining methyl function group produces CH2 or CH3, which inturn react with H to form CH4. Methane evolves earlier than

H2, and its release also ends earlier than H 2. The involved major

reactions are:

HH 2H (3)

RCH2R RR + CH2 (4)

CH2 +2H CH4 (5)

The decomposition of oxygen heterocyclic ring, both ether and

quinone in coal provides the major source of CO 2, while CO would

be mainly from the cracking of aliphatic matters, and some weak

bonds of aromatic and carboxylic groups. The cracked carboxylic

groups can react with O to form CO2 as well.


5/10


Correlating the preceding reaction pathways with Fig. 4 one can

see that the side-chains are easier to break to make CH4 release

earlier of pyrolysis reaction. The generated H would prefer to react

with some other active radials, causing the H2 release to appear

at high temperatures. Ether is easier to crack and CO2 can thus be

formed almost at the start of the pyrolysis reaction. Butthe decom-

position of quinine requires rather severe conditions so that the

CO2 concentration is low and its release preserves for long time

in pyrolysis. Cracking aliphatic matters would occur at interme-

diate temperatures, justifying the result in Fig. 4 that for CO and

hydrocarbons their releasing sequence and concentration values

are between those of CH4 and H2.

For the particle bed height of 350 mm Fig. 5 displays further

the yields of pyrolysis gas and oil varying with the reaction time.

As expected, the yields of pyrolysis oil and gas increased rapidly

with time in a period after feeding the coal sample into the reactor.

The oil yield reached its maximal value at about 200 s, while the

gas yield exhibited only a small increment (less than 20% of total

gas production) in the period from 200s to 350 s. The latter com-

plies well with the data in Fig. 4 where the release of pyrolysis gas

reached almost the end at the reaction time of 200s. Consequently,

from the viewpoint of maximizing the pyrolysis liquid product, a

reaction time of 3 minwouldbe enough. After 3 min, thecoal pyrol-

ysis is still in progress but the generated product is mainly gasthrough the reactions of polymerization or cracking of carbona-

ceous matters. Hence, the pyrolysis oil is produced mainly at the

initial stage of pyrolysis, while the later period of pyrolysis mainly

involves the reactions of carbonaceous residues. Fig. 5 clarifies also

that for the tested coal at the examined temperature of 973 K and

in N2 atmosphere, the tar yield with respect to coal mass of dry

ash-free basis reached about 12.3wt.%, while the corresponding

gas yield was 0.09L/g.

The time of 3 min was also proved to be sufficient for realizing

the maximal oil yield in the pyrolysis at the tested lower particle

bedheight of85 mm(referringfurtherto theinsettable comparison

in Fig. 5).

3.2. Further insight into the time-series behavior

The time-series pyrolysis gas release behavior is further anal-

ysed for the tested two particle bed heights in the reactor. One can

see from Fig. 4 that the general features were similar, but varying

the height affected the time reaching the gas concentration peaks.

The higher particle bed height caused the concentration peaks to

come earlier. Taking CH4 and CO as the examples, the time corre-

sponding to their concentration peaks were about 85s and110 s for

the particle bed heights of 350 and 85 mm, respectively. The result

complies with the fact that more hot particles in the higher bedcan

enhance the interaction between the hot bed material and the coal

sampleparticles to lead to a quicker rise of thecoal particle temper-

ature. This ensures consequently a quick pyrolysis in the reactor.

In practice, the particle bed height is generally decided by the per-mitted pressure drop through the reactor. According to Fig. 4, one

can suggest that theparticle bedheightshould be above 350 mm to

balance the reaction kinetics (heating rate) and the pressure drop

conditions.

The inset table in Fig. 5 compares the yields of pyrolysis gas

and oil realized at the two tested particle bed heights of 85 and

350mm at the reaction time of 190 s that led to the maximal oil

yields. The lower particle bed height caused the lower yields for

both gas and oil, indicating that the result was mainly related to

the quicker heating for coal particles in the 350-mm particle bed.

Although the higherfreeboard for85-mm particle bedwouldcause

deeper secondary cracking of tar to lower the oil yield and elevate

the gas yield [14], this effect is not dominant because the gas yield

was actually lower for the tested lower particle bed height.

Fig. 7. Effect of temperature on yields of pyrolysis oil and gas (atmosphere: N 2,

reaction time: 180 s).

Fig. 6 characterizes the pyrolysis oil composition with the TG-

FTIR spectra of the liquid products recovered at different reaction

time. Coal consists of many kinds of functional groups, and the

decomposition of these functional groups in pyrolysis [15] gener-

ates the pyrolysis gas and oil products. Similar to the gas product

characterized in Fig. 5, the composition of pyrolysis oil should alsovary with reaction time. Fig. 6 compares the FTIR spectra of the

pyrolysis oils collected at the reaction time of 25 s and 180 s. The

abscissa in Fig. 6 refers to the temperature of TG, and the anal-

ysed composition species include aliphatic hydrocarbons (CH3,

CH2, CH), single-ring aromatics (C C), carboxylic acids (C O),

phenol (OH (a)), and alcohol (OH (b)). The identified absorption

peaks in the FTIR spectra were shown in Table 2. The compared

pyrolysis oils had the similar FTIR spectra. There are two peaks for

aliphatic hydrocarbons and single-ring aromatics, while only one

peak appeared for the carboxylic compounds.

The FTIR data show that the oil from the pyrolysis for 180 s

containedmore aliphatichydrocarbonsand phenols butless single-

ring aromatic chemicals, carboxylic acids and alcoholic hydroxyl

compounds than the oil from the pyrolysis for 25 s. In pyrolysis,the various gaseous and liquid compounds come directly from the

breakage of side chains of large molecules and aromatics, and vary

with the second reactions occurring to the primary species. With

the longer reaction time of 180 s, the more primary volatiles were

surely released but for the tested 25 s the volatile release had to

be in process. Furthermore, the tested lignite is rich in aliphatic

hydrocarbons [16]. Considering all of these, one can believe that

in Fig. 6 the chemical species with higher contents for the pyroly-

sis by 180 s were generated mainly through the higher production

of the primary volatiles in the longer reaction time, while those

with the lower contents for the 180-s pyrolysis referred actually

to the obvious occurrence of the secondary reactions of those pri-

mary volatile species. This should be the case especially for the

carboxylic acids andhydroxide compoundsthat areeasy to decom-pose. Meanwhile, the higher production of aliphatic species for

lignite pyrolysis in the longer reaction time can also lower the

contents of the other chemical species containing in the pyrolysis

oil. Nonetheless, more fundamental studies are definitely needed

in order identify further the real causes for the results shown in

Fig. 6.

3.3. Pyrolysis temperature optimization

Reaction temperature is the critical parameter affecting the

pyrolysis product distribution and many studies have been con-

ducted to clarify the effect of temperature on the yields of gas and

liquid products in pyrolysis [5,17,18]. Fig. 7 shows how the yields

of pyrolysis oil and gas in the tested fluidized bed reactor varied


6/10


Fig. 8. TG-FTIR spectra of pyrolysis oil produced at different temperatures.

with the pyrolysis temperature. In the tested temperature range,

the pyrolysis gas yield increased continuously with the rise in the

temperature, whereas there was a peak yield forthe pyrolysis oilat

about 923 K. This temperature leading to themaximal oilyieldcon-

sisted with the resultof Cui et al. [5] who performed flash pyrolysis

of lignite in a fast-entrained bed reactor.

The appearance of a peak pyrolysis oil yield with varying the

reaction temperature shows in fact the competition between deep-

ening the pyrolysisreaction and enhancing the secondary reactions

such as cracking andreforming of the pyrolysis oil during elevating

the temperature. The former reaction is dominant at lower tem-

peratures to allow gradually increased pyrolysis oil production,

whilethe latter turnsto be overwhelming at highertemperaturesto

reducethe oilyield with raising thetemperature.Thesetwo types of

reactions all increase the gas production, causing thus the gas yield

to increase gradually from low to high temperatures. As one of the

major secondary reactions, the cracking of the pyrolysis oil would

occurfirst to the naphthenichydrocarbon andmacromoleculepolar

aromatics, andthen to thering-opening foraromatic compoundsat

higher temperatures. Accompanying the cracking, the polymeriza-

tion reaction is likelyto occur to increase thesemi-coke production

and to cause more heavy oil components. From the viewpoint of

achieving high light pyrolysis oil production, the coal pyrolysis

should thus be at temperatures below but close to the one leading

to the maximal pyrolysis oil yield, such as around 873 K according

to Fig. 7.

Fig. 8 shows theFTIRspectrameasuredvia TG-FTIR forthe pyrol-

ysisoils obtainedat two differentpyrolysis temperatures,823 K and


7/10


Fig. 9. Effect of reaction atmosphere on yield of pyrolysis oil (temperature: 873K,

reaction time: 180 s).

1023 K. Comparatively, the pyrolysis oilfrom the test at 1023K wascharacterized with obviously reduced spectrum intensity for the

plotted functional groups including hydrocarbons, phenols, alco-

holic hydroxide and single-ring aromatics. This further verifies that

coal pyrolysis for high-quality pyrolysis oil should not be at high

temperatures. Furthermore, Fig. 8 displays multiple peaks in the

FTIR spectra for single-ring chemicals, carboxylic acids, phenol,

and alcoholic hydroxyl group compounds, showing that there are

possibly different sorts of molecules for one kind of chemicals.

In summary, the pyrolysis temperature determines not only

the pyrolysis oil yield but also the oil composition. For the testing

system of this work, the pyrolysis oil yield reached its maximum

at 873923 K, meanwhile somehow reasonable yields of aliphatic

hydrocarbons and cyclane hydro-aromatics were also ensured. The

rather higher temperatures caused not only the lower pyrolysis oilyield but also the reduction in the contents of the valuable species

in the oil product (due to the deep secondary reactions occur-

ring to the primary pyrolysis oil). Therefore, in terms of producing

high-quality pyrolysis oil, it is critical to control the temperature at

reasonably low values that can ensure both high oil yield and high

oil quality.

3.4. Product distribution in varied reaction atmospheres

In the coal topping process illustrated in Fig. 1, coal pyrolysis is

likely to occur in pyrolysis gas atmosphere. Thus, a series of tests

were conductedto clarify theinfluences of various gascomponents

containingin the pyrolysis gas on pyrolysis behaviors.Figs.9and10

show the variation of pyrolysis oil yield with the changes of theatmospheric gas composition. For all the tests, the reaction time or

the time collecting pyrolysis oil after feeding coal sample into the

reactor was 180s and the temperature was fixed at 873 K.

Treating N2 as the basic atmosphere (see the inset table),

Fig. 9 demonstrates that adding H2 and CO2 into the atmosphere

decreased the pyrolysis oil yield, whereas further inclusion of CO

andCH4 intothe atmospheric gas increased the oil yieldconversely.

Fig. 10 replots the experimental data through correlating the oil

yield and the fraction of (H2 + CO2) in the atmosphere. It is clari-

fied that raising the contents of both these components decreased

monotonically the pyrolysis oilyield. By noting that the four datum

points on the left side were obtained through raising the H2 con-

tent and the other three points on the right side were from further

adding CO2 into the atmosphere, we can believe that the specific

Fig. 10. Influence of H2 and CO2 contents in the reaction atmosphere on pyrolysis

oil yield (temperature: 873K, reaction time: 180 s).

decrease in the oil yield with raising the gas content (i.e., the slope

of the two lines in Fig. 10) was almost the same for both H2 and

CO2.Most of the above-mentioned results are consistent with the

literature studies. Decreasing the pyrolysis oil yield with adding

CO2 into he atmosphere was reported by Cui et al. [18] for coal

pyrolysis in an entrained flow reactor, while increasing the volatile

production with including CH4 into the reaction atmosphere was

reported by Gao et al. [19] in their fixed bed coal pyrolysis tests

at 400750 C. According to Figs. 9 and 10, the presence of CO in

the atmosphere would enhance the formation of pyrolysis oil, but

there is no literature report yet regarding this influence.

As for the effects of H2, completely opposite reports are avail-

able in the literature. Several research groups [2025] reported

an obvious increase of the pyrolysis oil yield with increasing the

H2 content in the reaction atmosphere, while others [14,26] found

that the inclusion of H2 decreased the pyrolysis oil yield, althoughit improved the quality of the pyrolysis oil. Table 3 compares the

major literature reports and working conditions (including reac-

tor type and heating rate) for investigating the influence of H2 on

pyrolysis oil yield. It can be seen that increasing the H 2 content in

the reaction atmosphere increased the pyrolysis oil yield in all the

reported fixed bed pyrolysis tests but decreased the yields for the

tests in entrained flow or drop tube furnace reactors. The distinc-

tive difference for these two groups of reactors is the heating rate

of the coal sampleparticles insidethe reactor. In the fixed bedreac-

tor, the heating rate is limited to 10K/min, while theheating rate in

the entrained flow reactor and drop tube furnace can be as high as

10002000 K/s. This fact suggests that the influence of H2 on pyrol-

ysis performance is closely related to the heating rate for the coal

sample. Because the heating rate of fluidized bed reactor adoptedin this study was up to 1000K/s, the obtained result complied with

the literature finding in the entrained flow and drop tube reactors.

Table 3

Results of literature studies on influence of H2 on pyrolysis oil yield.

Reactor Heating rate Pyrolysis oil yield Ref. No.

Swept fixed bed reactor 3K/min [23]

Fixed bed reactor 5 K/min [22]

Fixed-bed reactor 10 K/min [21,24]

Fixed-bed reactor 150 K/min [19]

Tube reactor 1000 K/s [14]

Continuous free-fall reactor 2000K/s [25]

Note: the arrows and mean increase and decrease of pyrolysis oil yield with

raising H2 content in the atmosphere, respectively.


8/10


Fig. 11. TG-FTIR spectra of pyrolysis oil produced in different reaction atmospheres.

Fundamentally, the influence of atmospheric gas on pyrolysis

behavior is subject to two opposite types of essential interactions

[2629]. The gas may provide radicals, such as methyl, dimethyl, H

and carboxyl, to enhance the stabilizationof coal-baseradicals gen-

erated in coal thermal cracking to increase the pyrolysis oil yield.

Meanwhile, the gas may also interact with the formed gaseous

products, especiallythe condensable species (forming pyrolysisoil)

which have relatively large and long-chain molecules via the reac-

tions of hydrogenation, reforming and gasification to lower the oil

yield. When the former reaction is overwhelming, the pyrolysis oil

would be more with the inclusion of a gas into the atmosphere.

Otherwise, the increase of the fraction of the gas in the atmo-

sphere should decrease the oil yield. According to Figs. 9 and 10,

it is thought that in the tested fluidized bed reactor (heating rate

1000K/s), the gas components of CO2 and H2 would mainly par-

ticipate in the pyrolysis oil hydrogenation and reforming reactions,

whereas the supply of additional free radicals for coal-radical stabi-

lization would be dominant for CO and CH4 present in the reaction

atmosphere. The fact is that both H2 and CO2 are good reactant for

reformation and gasification, while CO and CH4 are easy to form

free radicals in comparison with thedirect reaction with oilspecies.

Notwithstanding, more studies are needed to clarify the mecha-

nisms of the preceding different effects of different gas species.

The gas G5in Fig. 9 simulates the pyrolysis gas composition. It is

shown that the compensative effects of (H2 + CO2) and (CO+CH4)

on pyrolysis made the pyrolysis oil yield in the simulated pyrolysis


9/10


gas atmosphere reach 13.21 wt.%, which was even higher than the

yield realized in N2 (about 12.9 wt.%). The result shows in fact that

using thepyrolysis gasas thereaction atmosphere would notaffect

greatly the pyrolysis oil yield in comparison with the pyrolysis in

N2.

3.5. Pyrolysis oil composition versus reaction atmospheres

The FTIR spectra in Fig. 11 display the composition features of

the pyrolysis oils made by the tests shown in Figs. 9 and 10. The

figure intends to clarify further how the gas atmosphere affects

the composition feature of the produced pyrolysis oil. The com-

pared atmospheres included G1, G2, G4 and G5. It is obvious that

the reaction atmosphere greatly affected the FTIR spectrum of the

pyrolysis oil in the TG-FTIR analysis. Including H2 into N2 atmo-

sphereobviouslydecreasedthe contents of aliphatic hydrocarbons,

phenol hydroxide, alcoholic hydroxide and aromatics in the pyrol-

ysis oil, while this meanwhile led to more acid species to present in

the liquid. Further addition of (CO + CO2) into (N2 + H2) influenced

the pyrolysis oil composition but the variation degree was much

smaller than changing the atmospheric gas from N2 to (N2 + H2).

Comparing to the (N2 + H2) atmosphere, the presence of (CO + CO2)

increased the formation of aliphatic hydrocarbons, phenols and

aromatics but little affected the contents of alcoholic hydroxideand carboxylic acid. By further including CH4 in the gas to make

the atmosphere simulate the pyrolysis gas, the composition of the

resulting oilbecame closer to that from thepyrolysis in N2 butwith

slightly higher contents for carboxylic acids, alcohols and phenols.

Meanwhile,many overall features of the FTIRspectrawere thesame

for all the cases tested. For example, at about 700 K there was a

peak absorbance for aliphatic hydrocarbons and at about 500K the

absorbance of carboxylic acids reached the maximum.

In summary, we can see that the pyrolysis in both N2 and the

simulated pyrolysis gas (H2 + CO2 + CO+ CH4) had not only very

close pyrolysis oil yields (Fig. 9) but also similar oil composition

features.Especially, the aliphatichydrocarbonsand aromaticsman-

ifested very similar FTIR spectra in the TG-FTIR analysis, while the

pyrolysis in the simulated pyrolysis gas has led to some slightlyhigher productions of the carboxylic acids, alcohols and phenols.

Therefore, using pyrolysis gas as the reaction atmosphere in coal

topping process is technical feasible, which even improves the

pyrolysis oil quality by causing more phenol production.

The preceding effects from Fig. 11 for the various atmospheric

gas components on the pyrolysis oil composition reflected as well

the influential essence clarified in the Section 3.4. Adding H2decreased the yields of all the major oil components, as a result

of its induced enhancement on the reformation and hydrogena-

tion of the oil species. The effects allowed by CO and CO2 in the

reaction atmosphere possibly compensated each other to make

the composition of FTIR spectra in Fig. 11 have no big differ-

ence between the pyrolysis oils produced in the atmospheres of

(N2 + H2) and (N2 + H2 + CO+ CO2). Further including CH4 into the(N2 + H2 + CO+ CO2) atmosphere obviously increased all the chem-

ical species characterized in Fig. 11. This, complying with the

literature reports of Liao et al. [30] andLiuet al. [31], demonstrated

actually a promotion effect of CH4 on the pyrolysis oil yield and

quality. The facilitated coal radial stabilization by the additional

free radicals generated from CH4 should be responsible for the for-

mation of the identified more light components including aliphatic

oil, phenols, alcohols and mono-aromatics in Fig. 11.

4. Conclusions

Pyrolysis of a kind of lignite in a laboratory fluidized bed reactor

under conditions simulating the so-called coal topping process in

a CFB boiler led to the following conclusions.

(1) The time to ensure the highest pyrolysis oil yield appeared to

be 180 s in a fluidized bed reactor at about 873 K. The TG-FTIR

analysis for the pyrolysis oil collected in different time period

from coal sample feeding clarified that the oil generated in the

early time of pyrolysis contained more aromatics, carboxylic

acids and alcohols butless aliphatic hydrocarbons and phenols.

Therefore, in order to get high-quality pyrolysis oil, the pyroly-

sis time should not be too short or too long, which was shown

to be between 50 and 180 s.

(2) In terms of pyrolysis oil yield and quality, the pyrolysis tem-

perature for the highest oil yield was shown to be about 873 K

(823923K), while the rather high temperature, such as at

1023K, dramatically decreased the contents of aliphatic hydro-

carbons, aromatics (mono-ring), phenols and alcohols in the

resulting pyrolysis oil. It appeared that too high temperature

would lead to more acid species. Therefore, the high-quality

pyrolysisoil characterizedby high contents of lightcomponents

(shortC chain andmono-aromatics) shouldbe generatedat rel-

atively low pyrolysistemperature, and 773900K should be the

recommended values.

(3) The effects H2 and CO2 in the reaction atmosphere on the coal

pyrolysis were proved to be mainly through their participation

in the secondary reactions of generated nascent pyrolysis oil,

including hydrogenation and reformation, to decrease the pro-duction of light components. Besides, the heating rate had a

great impact on the effect of H2 for coal pyrolysis. Both CO and

CH4 affected the pyrolysis principally via providing free radi-

cals to stabilize the coal radicals generated in coal molecular

cracking and breakage. Consequently, this increased the pyrol-

ysis oil yield and the contents of light components including

aliphatichydrocarbons,mono-aromatics, alcohols and phenols.

It is interesting to note that the pyrolysis in atmospheres of

N2 and the simulated pyrolysis gas (H2 + CO2 + CO+ CH4) had

very close pyrolysis oil yields, whereas the resulting oil in

the latter case contained slightly more carboxylic acids, alco-

hols and phenols. The contents of aliphatic hydrocarbons and

aromatics are equivalent for both reaction atmospheres. Con-

sequently, the use of pyrolysis gas as the reaction atmospherein the fluidized bed pyrolytic topping process is technical fea-

sible, which even allows slight upgrading of the pyrolysis oil

product.

This article also demonstrated that the different pyrolysis

gas species have different gas evolution characteristics, and

both CO2 and H2 had similar specific effect on decreasing the

pyrolysis oil yields with raising their contents in the reaction

atmosphere.

Acknowledgements

The authors are grateful to the financial support from National

Natural Science Foundation of China(No. 20776144), NationalBasicResearch Project of China (No. 2011CB201304) and National Key

Technology R&D Project of China (2009BAC64B05).

References

[1] X.B. Lin, G.B. Zhao, S.H. Wu, J. Yao, M. Kuxmk, et al., Mixing behavior betweencoal and hot circulation ash in down flow of coal topping process, J. Coal Conv.Prog. Chem. 30 (2007) 3641, Chinese.

[2] J.Z. Yao, M. Kwauk, A new process of coal topping for extracting liquid fuels, J.Process Chem. 7 (1995) 206210, Chinese.

[3] G.W. Xu, Z.E. Liu, S.Q. Gao, Value-added comprehensive processing of coalwith topping: strategy, technology and progress, in: The 4th Joint China/JapanChemical Engineering Symposium, China, 2007, pp. 2131, 30, Chinese.

[4] J.G.Wang,X.S. Lu,J.Z. Yao,Totaldistributionand liquid compositionof productsfrom coal topping process in a downer Reactor, Chin. J. Process Eng. 5 (2005)

241245, Chinese.


10/10


[5] L.J.Cui, W.G.Lin, J.Z.Yao, Influencesof temperatureand coalparticlesizeon theflash pyrolysis of coal in a fast-entrained bed, J. Chem. Res. Chin. U 22 (2006)103110, Chinese.

[6] L. Wang, M. Qin, H. Wang, Circulating ratio of heat carrier in coal toppingtechnological system, J. Energy Conserv. Technol. 24 (2006) 127130, Chinese.

[7] J.C.Bi, Development of cogenerationtechnologyintegratingpyrolysisand com-bustion in circulating fluidized bed, in: Presented in 2007s Symposium onComprehensive Utilization of Coal of Chinese Academy of Sciences, Beijing,2007, Chinese.

[8] S.F. Wen, Analysis on Fourier Transform Infrared Spectroscopy, second ed.,Chemical Industry Publisher, Beijing, 2010, Chinese.

[9] P.R. Solomon, D.G. Hamblen, R.M. Carangelo, et al., General model of coaldevolatilization, J. Energy Fuels 2 (1988) 405422.

[10] P.R. Solomon, M.A. Serio, E.M. Suuberg, Coal pyrolysis: experiments, kineticrates and mechanisms, J. Prog. Energy Combust. Sci. 18 (1992) 133220.

[11] X.D. Zhu, Z.B. Zhu, C.J. Han, Fundamental study of coal pyrolysis-functionalgroup and pyrolysisproducts, J. EastChina Univ. Sci. Technol.26 (2000) 1417,Chinese.

[12] H.M. Yang, L.L. Meng, M.J. Wang, et al., Effects of reaction atmospheres on theformation of gaseous productsduring coals pyrolysis, J. Taiyuan Univ. Technol.41 (2010) 338341, Chinese.

[13] J. Du, J.H. Wang, Y.P. Cui, et al., Forming and releasing of gaseous products ofcoal during pyrolysis in western China, J. China Univ. Ming Technol. 37 (2008)694698, Chinese.

[14] W.C. Xu,K. Matsuoka, H. Akiho,et al., High pressurehydropyrolysis of coals byusing a continuous free-fall reactor, J. Fuel 82 (2003) 677685.

[15] P.R. Solomon, D.G. Hamblen, in: R.H. Schlosberg (Ed.), Chemistry of Coal Con-version, Plenum Press, New York and London, 1985, pp. 121251.

[16] R.L. Yan, G.Y. Qian, Molecular structure of coal and gases produced by coaloxidation, J. Coal 20 (1995) 5864, Chinese.

[17] P. Wang, F. Wen, X.P. Bu, Study on the pyrolysis characteristics of coal, J. CoalConvers. 28 (2005) 813, Chinese.

[18] Y.P. Cui, L.L. Qin, J. Du Juan, et al., Products distribution and its influencingfactors for coal pyrolysis, J. Coal Chem. Ind. 129 (2007) 1015, Chinese.

[19] M.B. Gao, J.M. Zhang, M. Luo, Study of weight loss characters and promotingfunction to sulfur release during lignite pyrolysis in methane, J. Coal Convers.28 (2005) 710, Chinese.

[20] P.F.Wang,L.J. Jin,J.H. Liu.Analysis of coaltar derived frompyrolysis at differentatmospheres, J. Fuel., in press.

[21] H.Q. Liao, B.Q.Li, B.J. Zhang, Desulfurization and denitrogenation incopyrolysisof coal with hydrogen-rich gases, J. Fuel Chem. Technol. 27 (1999) 268272,Chinese.

[22] A. Ariunaa, B.Q. Li, W. Li, et al., Coal pyrolysis under synthesis gas, hydrogenand nitrogen, J. Fuel Chem. Technol. 35 (2007) 14.

[23] H.Q. Liao, C.G. Sun, B.Q. Li, Coal pyrolysis under hydrogen rich gases, J. Fuel

Chem. Technol. 26 (1998) 114118, Chinese.[24] M. Canel, Z. Misirlioglu, A. Sinag, Hydropyrolysis of a Turkish lignite (Tunc-

bilek) andeffect of temperatureand pressureon product distribution,J. EnergyConvers. Manage. 46 (2005) 21852197.

[25] R. Cypres, S. Furfari, Fixed-bed pyrolysis of coal under hydrogen pressure atlow heating rates, J. Fuel 60 (1981) 768778.

[26] H. Yabe, T. Kawamura, H. Kozuru, et al., Development of coal partial hydropy-rolysis process, J. Nippon Steel Tech. Rep. 382 (2005) 815.

[27] J.B. Howard, in: M.A Elliott (Ed.), Chemistry of Coal Utilization, John Wiley &Sons, New York, 1981, pp. 665784, 2nd suppl. vol.

[28] X.L. Fan, W. Zhang, Z.J. Zhou, Effects of pyrolysis pressure and atmosphere ongasificationreactivity of Shenfu char, J. FuelChem.Technol. 23(2005)530534,Chinese.

[29] N.O. Egiebor, M.R. Gray, Evidence for methane reactivity during coal pyrolysisand liquefaction, J. Fuel 69 (1990) 12761282.

[30] H.Q. Liao, W. Li, C.G. Sun, et al., Analysis on how to improve the pyrolysis oilyield during hydropyrolysis of coal, J. Chem. Ind. Eng. Prog. 6 (1996) 1923,Chinese.

[31] J.H.Liu, L.J.Jin, P.F.Wang,et al.,Research on integrated processof coalpyrolysis

withCO2 reformingof methaneto oil,in: InvitedPaperCollectionof 5thChineseNationalChemical andBiochemicalEngineeringAnnual Meeting,Beijing, 2008,pp. 120126, Chinese.

akiŞkan yatakli reaktÖrde kÖmÜr pİrolİzİ

Documents