Fuel 216 (2018) 810–817

Contents lists available at ScienceDirect

Fuel
journal homepage: www.elsevier.com/locate/fuel

Full Length Article

Gasification of dried sewage sludge using an innovative three-stage gasifier: T

Clean and H2-rich gas production using condensers as the only secondary tar
removal apparatus
⁎
Young-Kon Choi, Ji-Ho Ko, Joo-Sik Kim
Department of Energy and Environmental System Engineering, University of Seoul, 163 Siripdaero, Dongdaemun-Gu, Seoul 130-743, Republic of Korea

A R T I C L E I N F O A B S T R A C T

Keywords: Air gasification of dried sewage sludge was performed in a new-type three-stage gasifier consisting of an auger, a
Dried sewage sludge fluidized bed, and a tar-cracking reactors. The study aimed to evaluate the possibility of gasification using only
Gasification condensers for the secondary tar removal method. The effects of Ni- and Fe-impregnated active carbons on the
Tar quality of producer gas were also investigated. Finally, to investigate the deactivation behavior of active carbon,
Producer gas
a spent active carbon obtained after about 4.3 h of gasification, was used again for approximately 3.7 h of
Active carbon
gasification. The active carbons used effectively reduced the contents of condensed and gaseous tars. Ni-im-
pregnated active carbon produced a gas with a high H2 content (26 vol%) and a low NH3 content (198 ppmv),
while Fe-impregnated active carbon produced a gas with a low H2S content (96 ppmv). The compositions of the
producer gases obtained solely using condensers were similar to those obtained using electrostatic precipitator.
During the total ∼8 h of gasification, active carbon could efficiently remove tar; however, its surface area and
total pore volume slowly decreased with time.

1. Introduction
Due to a rise in global energy demand, attention is focused world-
wide on producing renewable energy from various biomass sources.
Gasification, a thermo-chemical conversion process, is a technique that
has been in use since 1839, when the wood gasifier was built, and has
recently been identified as a solution for producing renewable energy
from biomass [1,2]. The gas produced from biomass gasification, called
producer gas or syngas, can be used as a fuel for heat or electricity
generation, or as a feedstock for chemical and fuel synthesis [3]. Bio-
mass gasification is a promising route for H2 production, because bio-
mass-derived H2 contributes to renewable energy production [4]. Im-
purities that evolve during biomass gasification can cause issues that
need to be resolved to allow commercialization of gasification pro-
cesses. For example, tar can result in mechanical problems, such as
blocking the pipelines of gasification processing or fouling processing
equipment, but it can also deactivate catalysts used during gasification
[5,6]. Additionally, tar in producer gas prevents its direct use in gas
engines or turbines. Milne and Evans proposed acceptable guideline
values for tar content in producer gas for devices, as follows: less than
5 mg/Nm3 for direct-fired industrial gas turbines, and 50–100 mg/Nm3
for internal combustion engines [7]. Tar removal approaches are often
divided into two methods: primary and secondary. The primary method
relies on tar removal inside gasifiers, using tar-cracking additives and/
or optimal tar-cracking parameters. The secondary method uses appa-
ratus after gasification, such as a scrubber or an electrostatic pre-
cipitator (EP). In addition to tar removal, the secondary method is also
effective in removing other impurities (H2S, NH3, and HCl) [8]. Cur-
rently, most biomass gasification processes that are running or being
tested are equipped with a scrubber and (or) an EP. This method,
however, poses significant problems when commercialization of a
biomass gasification process is considered, because they limit the eco-
nomic efficiency of the process [9]. Another secondary method is to use
catalysts in tar-cracking or reforming reactors. Wang et al. used Ni/char
catalysts, and found that the catalysts led to the removal of over 97% of
tar at 800 °C and increased H2 in syngas [10]. These catalysts were also
tested for removing tar inside gasifiers [11–13]. Garcia et al. in-
vestigated the effect of a Ni-Al catalyst as a fluidized bed material, and
concluded that the Ni-Al catalyst enhanced H2 production and de-
creased methane content in producer gas [14]. Virginie et al. examined
the effect of Fe/olivine, which is inexpensive and non-toxic, on tar re-
moval in a dual fluidized bed biomass gasification process, and con-
cluded that it acted as a catalyst for reforming tar and hydrocarbons
[15]. Ko et al. reported that Fe is effective as a sorbent of H2S between

⁎
Corresponding author.
E-mail address: joosik@uos.ac.kr (J.-S. Kim).

https://doi.org/10.1016/j.fuel.2017.12.068
Received 9 October 2017; Received in revised form 4 December 2017; Accepted 17 December 2017
0016-2361/ © 2017 Elsevier Ltd. All rights reserved.
Y.-K. Choi et al. Fuel 216 (2018) 810–817

Nomenclature EP Electrostatic precipitator

ER Equivalence ratio
List of abbreviations GC-TCD Gas chromatography-thermal conductivity detector
GC-FID Gas chromatography-flame ionization detector
BET Brunauer–Emmett–Teller ICP-AES Inductively coupled plasma atomic emission spectrometry
CCE Carbon conversion efficiency TRE Tar removal efficiency
CGE Cold gas efficiency UOS University of Seoul
DSS Dried sewage sludge

the temperatures of 500–700 °C [16]. Most catalysts used as primary
and secondary methods, however, suffered from deactivation by coke,
sulfur, and chlorine [17,18]. Using modified gasifiers as a primary
method is very promising, because this does not rely on expensive
equipment, like EPs and catalysts. Some approaches reported that
modified gasifiers could help overcome the problems associated with
tar contaminants, resulting in enhanced H2-rich syngas yield and pro-
cess efficiency [19,20].
Our group has developed a staged gasifier (the UOS two-stage ga-
sifier) consisting of a fluidized bed gasifier and fixed bed reactor serially
connected, and we conducted a series of gasification of dried sewage
sludge (DSS) [21–24]. The two-stage gasifier used active carbon as a tar
removal additive, and showed excellent tar removal efficiency with
high H2 production. Recently, a new type of staged gasifier, called the
UOS three-stage gasifier, consisting of three serially connected reactors
(auger, fluidized bed, and tar cracking reactors), was developed to re-
move tar more efficiently than the UOS two-stage gasifier. This paper
reports for the first time the experimental results of the three-stage DSS
gasification, solely using condensers as the primary tar removal method
without using any typical expensive tar removal apparatus, such as an
EP or a scrubber [25]. This paper also reports the effects of Fe and Ni-
impregnated active carbons in the three-stage gasification on the im-
purity removal (tar, NH3, and H2S), and on the H2 production. In ad-
dition, Fe-impregnated active carbon was firstly tested in this paper to
reduce the H2S content in producer gas, and its effectiveness was also
provided in detail. Finally, this paper provides results on a long-term
gasification (total ∼8 h) without any EP and scrubber to observe the
deactivation behavior of active carbon and producer gas development,
which was the longest operation with DSS in the UOS staged gasifier.
2. Experimental
2.1. Feed material and active carbon
In the experiments, DSS obtained from anaerobic digestion supplied
by a wastewater treatment facility in Korea was used. The DSS was first
crushed by a disintegrator and sieved to obtain homogenous fractions
with diameters ranging between 0.6 and 3.35 mm. The main char-
acteristics of the feed material are presented in Table 1.
Table 1 shows that the DSS contained high nitrogen (∼5 wt%) and
sulfur (∼2 wt%) contents. Nitrogen and sulfur in feed material are the
precursors for NH3 and H2S liberated during gasification. The DSS had a
high ash content (∼40 wt%), with Fe, Ca, Mg, and Al as the main metal
components. The lower heating value of DSS determined using a bomb
calorimeter (Model 6100, Parr) was about 13 MJ/kg. Silica sand was
used as the bed material in the fluidized bed gasifier, which had a
diameter between 150 and 300 μm. A coal-based virgin active carbon
with a size of 1–2 mm and a surface area greater than 1030 m2/g was
used as a tar-cracking additive in the three-stage gasification process.
Along with this, metal-impregnated (Ni and Fe) active carbons were
also used. The surface areas of the Ni- and Fe-impregnated active car-
bons, with metal content amounting to ∼9 wt%, were approximately
960–980 m2/g. Metal impregnation was conducted as follows: first, a
metal (either Ni or Fe) ion solution was mixed with a coal-based active
carbon and stirred. During stirring, the mixture was heated for ∼4 h to
remove any water content. After heating, the remainder was dried
overnight at 90 °C, and then calcined at 400 °C to activate it and remove
impurities.
2.2. The UOS three-stage gasification process
A diagram of the three-stage gasification process is presented in
Fig. 1.
The three-stage gasifier consists of an auger, a fluidized bed, and
tar-cracking reactors, connected serially. The fluidized bed and tar-
cracking reactors are separated by a perforated distributor with 1 mm
pores. The auger reactor, which is absent from the UOS two-stage ga-
sifier, has a length of 850 mm and an inner diameter of 32 mm and is
located in front of the fluidized bed reactor. The auger reactor plays an
important role to reduce the tar content in producer gas: it generates a
huge amount of tar in advance of the main gasification in the fluidized
bed reactor. A portion of the tar generated in the auger reactor can be
destroyed through thermal and oxidative thermal decomposition inside
the auger reactor. The undestroyed tar leaving the auger reactor can
have a long time to react with air used as the gasifying agent inside the
fluidized bed reactor, compared to tar directly generated inside a ga-
sifier. The fluidized bed reactor of the three-stage gasifiers, which is the
main gasifier, has a height of 390 mm and an inner diameter of
110 mm. Tar can also be produced inside the fluidized bed gasifier by
the gasification of rest feed material from the auger reactor and it then
enters the tar cracking reactor, which has a height of 410 mm and an
inner diameter of 160 mm. The active carbon in the tar cracking reactor
absorbs and cracks tar. After gasification in the three-stage gasification
process, producer gas can be cleaned by condensers or an EP. Details of
the process elements and the operational procedure for the three-stage
gasification process have been previously published [26,27].
2.3. Experimental conditions
In all the experiments, the feed rate was 1 kg DSS/h, and gasification
time ranged from ∼1 to 4.3 h. The amount of silica sand used as the
fluidizing bed material was 2.5 kg. Preheated air (∼550 °C) was used as
the gasifying agent, and the air flow rates for all of the experiments
were ∼17 Nl/min. The amount of active carbon used, whether it was
the virgin or metal impregnated active carbons, was 1 kg. The contact
Table 1
Compositions of dried sewage sludge.
Dried sewage sludge
Proximate analysisa (wt%) Ultimate analysisb (wt%)
Moisture 7.07 ± 0.31 Carbon 31.32 ± 0.11
Volatile matter 51.53 ± 1.39 Hydrogen 4.56 ± 0.08
Fixed carbon 3.99 ± 0.28 Nitrogen 4.72 ± 0.02
Ash 37.41 ± 1.02 Oxygenc 20.72 ± 0.31
Sulfur 1.27 ± 0.01
Lower heating valuec (MJ/kg) 13.2
a
ASTM E871-82, ASTM E872-82, ASTM D1102-84.
b
dry basis.
c
calculated by difference.

811
Y.-K. Choi et al. Fuel 216 (2018) 810–817

Fig. 1. Diagram of the three-stage gasification process (The UOS three-stage gasification process).

time between the producer gas and active carbon bed in the tar- efficiency (TRE), which was an indicator of the efficiency of the three-
cracking reactor was ∼1 s (calculated as the voidage of the active stage gasification process for tar removal, was also defined using total
carbon bed (0.55) divided by the total volumetric flow of producer gas tar. The three parameters were determined using the following for-
at the reaction temperature). In this study, the equivalence ratio (ER) mulas:
was defined as the air/fuel weight ratio used in gasification divided by
Total carbon(C1–C5)in producer gas
the air/fuel weight ratio for stoichiometric combustion: it was main- CCE (%) = × 100
Total carbon in DSS (1)
tained at ∼0.3 for all the experiments. Throughout the experiments,
the temperatures of the auger, fluidized bed, and tar-cracking reactors Caloric value of producer gas
CGE (%) = × 100
were maintained at ∼650, ∼815, and ∼815 °C, respectively. The Caloric value of DSS (2)
specific experimental conditions are shown in Table 2.
Total tar with AC (g)
Run 1 and 2 were performed without and with virgin active carbon, TRE (%) = 1− × 100
respectively, to determine the role of active carbon in tar removal. From Total tar without AC (g) (3)
Runs 3 to 7, gasification was performed with condensers as the only
secondary tar-removal method. Comparisons between Runs 2 and 3, 2.4. Analysis of products
which were conducted with virgin active carbon, would reveal the
difference between gasification with and without an EP. In Runs 4 and All the gasification products were divided into four groups: pro-
5, Ni- and Fe-impregnated active carbons were used in the tar-cracking ducer gas, char, and condensate liquid collected in the condensers and
reactor. Run 6 was conducted for ∼4.3 h to study the development of EP, and tar deposited on the process. Producer gas was analyzed using a
producer gas and the deactivation behavior of virgin active carbon with GC–TCD and GC–FID (7890A, Agilent Instruments) with argon as the
gasification time. Run 7 was carried out using the spent active carbon carrier gas. The types of columns used for GC–TCD and GC–FID were
from Run 6 to determine the reusability of active carbon, and these Carboxen 1000 and HP–plot Al2O3/KCL, respectively. In this study, the
results were compared with those of Run 3, which was conducted for tar in producer gas was identified as molecules larger than benzene,
1 h. To evaluate the efficiency of each run in the three-stage gasifica- according to the International Energy Agency’s guidelines [28]. Hence,
tion, the cold gas efficiency (CGE) and carbon conversion efficiency the tar content in producer gas was determined from the ratio between
(CCE) of each run were compared. The total amount of tar condensed the sum of areas of all the peaks after the benzene peak and the sum of
from the gasification process was denoted as total tar. Tar removal peaks of all components. The NH3 and H2S contents in producer gas

Table 2
Experimental conditions.

Experimental conditions Run 1 Run 2 Run 3 Run 4 Run 5 Run 6 Run 7

Auger reactor temperature (°C) 642 644 645 645 653 661 651
Fluidized bed reactor temperature (°C) 816 813 817 815 817 814 816
Tar cracking reactor temperature (°C) 813 815 813 817 815 816 819
Gasification time (min) 61 60 61 60 59 260 220
Electrostatic precipitator On On Off Off Off Off Off
Additive type (amount) None (–) Virgin AC (1 kg) Virgin AC (1 kg) Ni/AC* (1 kg) Fe/AC* (1 kg) Virgin AC (1 kg) Spent AC* (1 kg)

AC: active carbon, Ni/AC*: Ni 9.1 wt%, Fe/AC*: Fe 8.9 wt%, Spent AC*: spent AC after Run 6.

812
Y.-K. Choi et al. Fuel 216 (2018) 810–817

were measured using a gas measurement analyzer (IR 9000, Rbr). The NH3 content in producer gas declined from 1794 (Run 1) to 893 ppmv
detection precision for the NH3 and H2S quantification was ± 5 ppmv. (Run 2) with the addition of active carbon. This could be due to the
Additionally, the surface areas of the active carbons were measured high NH3 absorptivity of active carbon by its oxygen functional groups
using a Brunauer–Emmett–Teller (BET) surface area analyzer. The [29]. In summary, the experiment with active carbon yielded a pro-
contents of the metals impregnated on active carbon were analyzed ducer gas with two times as much H2, 10 times less tar, and two times
using inductively coupled plasma atomic emission spectroscopy (ICP- less NH3 content than that from the experiment without active carbon.
AES) (ICPE-9000, Shimadzu).
3.1.2. Ni- and Fe- impregnated active carbons
3. Results and discussion Table 3 presents the effects of Ni and Fe impregnated active carbons
on producer gas quality. First, Ni-based catalysts are reported to effi-
3.1. Effects of virgin and metal impregnated active carbons ciently remove tar by cracking, steam reforming reactions, and dry
reforming reactions [10,31,32], so enhanced the H2 production and
The experimental results of this study are presented in Table 3. decreased tar. Run 4 confirmed this expectation: an increase in H2
All the producer gases obtained using the active carbons had very (28.1 vol%) compared to Run 3 (25.2 vol%). Compared to the UOS two-
low impurity contents (tar, NH3, and H2S). Heating values of the pro- stage gasification (Run 6: 8.9 wt% Ni on active carbon) [24], the H2
ducer gases obtained using the active carbons were maintained at content of Run 4 in this study was slightly decreased mainly because
∼5 MJ/Nm3 due to the low hydrocarbon contents of producer gases. sand was applied in the present work as the fluidized bed material in-
CGE, CCE, and TRE ranged from 68 to 89%, 76 to 86%, and 67 to 93%, stead of olivine. The producer gas from Run 4 of the present work
respectively. contained a low level of tar (95 mg/Nm3), and the TRE of Run 4 was
much higher (91%) than that of Run 3. Additionally, the producer gas
3.1.1. Virgin active carbon yield, CGE, and CCE of Run 4 increased to 75.1 wt%, 89%, and 85%,
For the removal of tar generated during gasification, many re- compared to Run 3. Nitrogen in DSS is converted to NH3 during gasi-
searchers have focused on the application of catalysts or additives from fication, and Ni is an efficient catalyst for the NH3 removal [10]. The
natural sources, due to their ready availability and low cost. Active Ni-impregnated active carbon used in this work (Run 4) reduced the
carbon is particularly advantageous, because it has high adsorption NH3 content, with the lowest values (198 ppmv on average) observed
selectivity for hydrocarbons [29]. Comparing Runs 1 and 2 in Table 3, from the experiments. Considering the low NH3 content (119 ppmv)
the addition of coal-based virgin active carbon brought about an in- from the UOS two-stage gasification (Run 6: 8.9 wt% Ni on active
crease in H2 from ∼13 (without active carbon) to ∼25 vol% and a carbon) [24], the introduction of Ni-impregnated active carbon in the
decrease in tar from 2544 (without active carbon) to 235 mg/Nm3 in UOS stage gasification appears to be a strong measure against the NH3
producer gas. According to a previous study [30], typical H2 contents of
producer gases from air gasification of biomass ranges between 5 and Table 3
16 vol%. In the experiments, the active carbon in the tar cracking zone Experiment results.
had a decisive effect on the removal of tar and H2 production. The
Composition (vol Run 1 Run 2 Run 3 Run 4 Run 5 Run 6 Run 7
impressive performance of tar removal by the active carbon is due to its %)
tar absorptivity, and tar cracking by reactions such as thermal or cat-
alytic tar cracking, water-gas reactions, and steam reforming: N2 51.76 47.81 46.86 44.12 47.85 48.13 50.37
CO2 14.45 11.28 11.50 9.36 10.72 11.43 11.90
Thermal or catalytic cracking: Cn Hm ⇌ C+ C x Hy + H2 (4) H2 13.42 25.02 25.21 28.13 26.33 23.65 20.94
CO 10.06 10.92 11.21 13.38 10.78 12.16 12.39
Carbon formation: Cn Hm ⇌ nC + (x/2)H2 (5) CH4 6.46 4.91 4.87 4.93 4.23 4.41 4.01
C2H2 0.15 N.D. 0.03 0.01 0.01 0.02 0.03
Steam reforming:Cn Hm + nH2 O⇌ nCO + ( n+ m/2)H2 (6) C2H4 2.18 0.02 0.21 0.06 0.06 0.12 0.21
C2H6 0.19 N.D. 0.03 N.D. N.D. 0.02 0.02
Dry reforming:Cn Hm + nCO2 ⇌ 2nCO + (m/2)H2 (7) C3 + C4 + C5 0.08 0.01 0.01 < 0.001 0.01 0.01 0.01
Benzene 1.18 0.02 0.06 0.01 0.01 0.04 0.06
CnHm hydrocarbons are tars (i.e., compounds with molecular masses > Benzene 0.07 0.01 0.01 < 0.001 < 0.001 0.01 0.05
larger than that of benzene), while CxHy represents hydrocarbons with LHV (MJ/Nm3) 6.52 5.31 5.51 5.90 5.22 5.13 5.04
carbon numbers smaller than those of CnHm. Product (wt%)
The above reactions remove tar and, at the same time, increase the Producer gas 67.0 73.3 72.8 75.1 74.3 70.8 70.5
H2 content in producer gas. The surface area of the spent active carbon Char 19.9 17.0 17.7 16.2 16.3 19.1 19.1
Condensate 10.8 9.3 9.2 8.5 9.3 9.5 9.6
obtained after Run 2 decreased from ∼1030 to 590 m2/g. The sig-
liquid
nificant drop in surface area could cause a negative effect on the tar Tar 2.3 0.4 0.4 0.2 0.2 0.6 0.8
removal and H2 production. Fig. 2 shows the development of producer
Process Efficiency (%)
gas from Run 2, along with gasification time (Run 2). Cold Gas 73.43 75.24 73.38 88.87 75.66 71.82 67.53
In Fig. 2, the tar content in producer gas gradually increased to Efficiency
∼300 mg/Nm3 from the beginning to a gasification time of 20 min, Carbon 86.24 78.02 81.08 84.60 75.79 76.78 76.37
which then stabilized from 20 min to the end of the gasification process, Conversion
Efficiency
but the H2 content of the producer gas from Run 2 was maintained at a
Tar removal Base 81.41 82.17 91.0 92.9 77.2 67.2
high level (∼25 vol%) throughout gasification. The increase in tar from Efficiency
the beginning to 20 min suggests a gradual blocking of active carbon
Impurities in producer gas
pores. During this period, tar still appeared to be cracked, constantly Tar content 2544 235 250 95 41 466 786
producing H2. After 20 min, majority of the active carbon pores, which (mg/Nm3)
mainly faced the up-flowing producer gas, were blocked by tar and NH3 content 1794 893 905 198 913 894 914
coke. During this time, the cracking of tar and coke on active carbon (ppmv)
H2S content – 724 735 698 96 787 1139
continued to take place by Eqs. (4)–(7), producing H2. The continuous
(ppmv)
cracking of tar and coke on active carbon suggests the regeneration of
active carbon. Detailed explanations on deactivation and regeneration N.D.: not detected (< 0.0001 vol%), –: beyond detection range (NH3: 3000 ppmv, H2S:
of active carbon have been published in previous studies [26,27]. The 2000 ppmv).

813
 Y.-K. Choi et al.
30
25
Producer gas composition (vol%)
20
15
10
0 0
0 10 20 30 40
Gasification time (min)
production in DSS gasification. The Ni catalyst, however, can be easily
deactivated by sulfur compounds like H2S. Many papers reported the
metal deactivation through sulfur poisoning [17,33–35]. The formation
of nickel sulfide, i.e., the deactivation of Ni, can be described as follows
[36]:
Nickel sulfide formation:Ni + H2 S⇌ Ni-S + H2
As shown in Fig. 3(a), which depicts the development of tar, NH3,
H2S, and NOx with time in Run 4, the NH3 content in producer gas was
maintained at relatively low levels (below 300 ppmv) compared to
Fig. 3(b), which contains the results of Run 5 (Fe). A noticeable increase
in tar in producer gas with time would indicate that Ni-impregnated
active carbon was unavailable, which could have been due to the de-
activation of Ni by a reaction with H2S and the blocking of active
carbon pores by tar and coke [17,36]. In order to reduce the degree of
deactivation of Ni-impregnated active carbon, the blocking of active
carbon pores by tar and coke at least should be minimized. In–situ re-
generation of active carbon using oxidants like air would lessen the
pore blocking of active carbon and therefore be one realizable method
against the deactivation of active carbon. In contrast to NH3, the H2S
content of the producer gas from Run 4 was maintained at a high level
throughout gasification.
Table 3 shows that gasification with Fe-impregnated active carbon
(Run 5) had a similar product distribution to Run 3 (virgin active
carbon). Tar cracking, however, was enhanced when Fe-impregnated
active carbon was used. The producer gas from Run 5 contained a very
low level of tar (41 mg/Nm3), and TRE for Run 5 was very high (92%).
The catalytic effect of Fe in tar cracking has also been described in other
work [37]. Fig. 3(b) shows a sharp increase in the tar content of pro-
ducer gas after 30 min, similar to the case of Ni-impregnated active
carbon. Table 3 shows the average H2S content in the producer gas from
Run 5 was very low, which was 96 ppmv on average. Fig. 3(b) shows
that the H2S content in the producer gas slowly increased with time, but
did not surpass 300 ppmv. The following reaction could have been re-
sponsible for the H2S capture.
Iron sulfide formation:Fe + H2 S⇌ FeS + H2
Fe-impregnated active carbon turned out to be ineffective in NH3
reduction. In summary, Ni- and Fe-impregnated active carbons used in
the three-stage gasifier sharply reduced tar content in producer gas and
the amount of condensed tar. Ni-impregnated active carbon was very
Fuel 216 (2018) 810–817
Fig. 2. Development of producer gas with gasification
time (Run 2).
300
Tar in producer gas (mg/Nm3)
H2
CO
CH4
200
CO2
Tar
100
50 60
effective in NH3 reduction, reducing its activity with time. Fe-im-
pregnated active carbon, however, showed its high activity for the H2S
reduction. In–situ regeneration of active carbon appears to be effective
in keeping the activity of the catalysts in a high level.
(8) 3.2. Gasification without EP
Secondary tar removal methods, such as EPs or scrubbers, are fairly
expensive, and the energy in the tar separated by the secondary
methods is usually lost. Comparisons between Run 2, carried out with
an EP, and Run 3, carried out without an EP, are provided in Table 3.
Changes in producer gas composition and product distribution between
the two runs are negligible. There was a small increase in the tar con-
tent of the producer gas from Run 3, but this appeared to be insignif-
icant. Overall, gasification without an EP did not have any negative
influence on producer gas quality for at least 1 h of gasification, in-
dicating that tar removal primarily occurred inside the three-stage ga-
sifier itself. Runs 4–6, which were also conducted without an EP, pro-
duced high quality producer gas.
3.3. Long-term gasification and reusability of active carbon
To determine the stability of the three-stage gasifier, especially in
terms of deactivation of active carbon, long-term gasification was
tested. At first, Run 6 was conducted using virgin active carbon, but the
gasification time was limited to ∼4.3 h due to ash accumulation in the
fluidized bed reactor. To test the activity of active carbon for a long-
term run, the spent active carbon obtained from Run 6 was used for Run
7, which lasted ∼3.7 h. Table 3 shows that the producer gas from Run 6
had, on average, a relatively low H2 content and high impurity content
compared to the producer gas from Run 3 (virgin active carbon, 1 h): 1)
the tar content in producer gas increased from 250 to 466 mg/Nm3, 2)
the TRE decreased from 82.2 to 77.2%, and 3) H2 production decreased
from 25.2 to 23.7 vol%. These results suggest that the activity of active
carbon decreased with time. Fig. 4 shows the development of producer
(9) gas with gasification time for Runs 6 and 7. In the figure, the producer
gas development of Run 7 is combined with that of Run 6 and gasifi-
cation time.
As shown in Fig. 4(a), a steady state for Run 6 was reached ∼50 min
after the run began, and no noticeable change in producer gas com-
position was observed after that. The producer gas composition of Run
814
 Y.-K. Choi et al. Fuel 216 (2018) 810–817

800 200 Fig. 3. Changes of impurity content with time: (a) Ni/
(a) active carbon (b) Fe/active carbon.

Tar in producer gas (mg/Nm3)

600 150
Impurity content (ppmv)

NH3
H2S
NOx
Tar
400 100

200 50

0 0
0 10 20 30 40 50 60

Gasification time (min)

1000 100
(b)

800 80

Tar in producer gas (mg/Nm3)

Impurity content (ppmv)

600 60
NH3
H2S
NOx
400 Tar 40

200 20

0 0
0 10 20 30 40 50 60

Gasification time (min)

6, until a gasification time of ∼ 1 h, was very similar to that of Run 3,
however after ∼1 h, the H2 content in producer gas slightly, but gra-
dually, decreased, indicating gradual deactivation of active carbon.
Considering the average H2 and tar content of the producer gas from
Run 3 (1 h), the active carbon of Run 6 appeared to play a successful
role in tar cracking until ∼4.3 h. The gas composition development of
Run 7 in Fig. 4(a) reveals that there was almost no change in producer
gas composition, except for a small decrease in H2 content when the
spent active carbon from Run 6 was used. Significant deterioration of
spent active carbon as an additive for tar removal was not observed.
The small decrease in H2 content from Run 7 suggests that repetitive
deactivation and regeneration of spent active carbon took place con-
tinuously, and the deactivation rate was slightly faster than the re-
generation rate. Repetition of an increase in tar content and stabiliza-
tion of tar content from 100 min to the end of gasification can be
observed in Fig. 4(b), which also shows that NH3 and NOx content was
stabilized throughout the long-term gasification time (total ∼8 h) while
H2S content increased. An increase in H2S content with time, ob-
servable in Fig. 4(b), could be a result of the deactivation of active
carbon. The meso- and micro-pores of active carbon have been reported
to be important for H2S adsorption [29,38].
The TRE value is another important indicator of the deactivation of
active carbon. The TRE of Run 7 was considerably lower (67.2%) than
Run 6 (77.2%). This indicates that the spent active carbon in Run 7 lost
its ability to absorb tar molecules to some extent, and an increased
portion of tar, compared to Run 6, passed through the active carbon bed
without being captured. Table 4 shows the textural properties of the
(spent) active carbons.
As shown in Table 4, there was a decrease in the surface area and
total pore volume of active carbon after 1 h of gasification (Run 3). This
trend was also observed for metal-impregnated active carbons. After
1 h, the rate of surface area and total pore volume decrease was re-
duced, which could be an evidence of the regeneration of active carbon.
The spent active carbon after Runs 6 and 7 still had considerably large

815
Y.-K. Choi et al. Fuel 216 (2018) 810–817

Fig. 4. Producer gas development with time (a) gas

composition (b) impurities.

Table 4 0.8
Textural properties of (spent) active carbons.
Fresh AC
Spent AC (Run3)
dVp /dlog(dp) (cm /mg/g)

BET analysis BET surface Total pore Average pore 0.6 Spent AC (Run6)
area (m2/g) volume (cm3/g) diameter (nm) Spent AC (Run7)
3

AC without metals
Virgin 1030.5 0.5345 1.6876 0.4
Spent AC after 590.3 0.3869 1.4878
∼1 h (Run 3)
Spent AC after 493.6 0.2911 1.2130
0.2
∼4.3 h (Run 6)
Spent AC after 392.7 0.2695 1.1876
∼8 h (Run 7)

Ni-impregnated AC 0.0
1 10 100
Virgin 958.6 0.5098 1.7040
Spent AC after 568.6 0.3528 1.4492 Pore diameter (dp ) (nm)
∼1 h (Run 4)
Fig. 5. Pore size distribution of active carbons.
Fe-impregnated AC
Virgin 976.1 0.5141 1.6926
Spent AC after 613.2 0.3717 1.5137 surface areas (493 and 392 m2/g) and total pore volumes (0.2911 and
∼1 h (Run 5) 0.2695 cm3/g). Fig. 5 shows the pore size distributions of virgin and
spent active carbons used in this work.
AC: active carbon.
As shown in Fig. 5, pores, especially micro-pores, on virgin active
carbon gradually reduced after gasification due to coke formation on
active carbon. Table 5 provides further evidence of coke formation on

816
Y.-K. Choi et al.
Table 5
Weight of active carbon after gasification.
Experiment Active carbon (g)
Before After
Run 3 1000 for all Runs 1038
Run 4 1031
Run 5 1026
Run 6 1209
active carbon, showing an increase in the weight of active carbon after
the experiments.
From the above results, it could be concluded that extensive con-
sumption of active carbon during gasification was unlikely to occur,
because the tar and coke on active carbon protected it from being
consumed. To summarize, deactivation of active carbon during gasifi-
cation in the three-stage gasifier is inevitable. When active carbon was
deactivated over time, the tar and H2S content of producer gas gradu-
ally increased, but NH3 content remained constant. The ability of active
carbon to capture condensable tar diminished with time, but was still
considerable. One method for regenerating active carbon would be
intermittent operation at a high ER to remove tar or coke.
4. Conclusions
Three-stage gasification of dried sewage sludge using only con-
densers yielded a producer gas similar in gas composition to those
obtained using an electrostatic precipitator. In this study, tar removal
occurred mostly inside the three-stage gasifier. The results could be
used to attempt to commercialize biomass gasification systems, because
the capital cost of an electrostatic precipitator is very high. The pro-
ducer gas generated using Ni-impregnated active carbon contained
26 vol% H2 at 198 ppmv NH3. Fe-impregnated active carbon strongly
reduced the H2S content (96 ppmv) of producer gas. Long-term gasifi-
cation revealed that surface area and the total pore volume of active
carbon decreased within 1 h after gasification. Thereafter, however, the
deactivation rate was diminished. The deactivation and regeneration of
active carbon took place simultaneously.
Acknowledgment
This work was supported by the New & Renewable Energy Core
Technology Program of the Korea Institute of Energy Technology
Evaluation and Planning (KETEP) granted financial resource from the
Ministry of Trade, Industry & Energy, Republic of Korea (No.
20153030091340).
References
[1] Laksmono N, Paraschiv M, Loubar K, Tazerout M. Biodiesel production from bio-
mass gasification tar via thermal/catalytic cracking. Fuel Process Technol
2013;106:776–83.
[2] Pacioni TR, Soares D, Di Domenico M, Rosa MF, Moreira RDFPM, José HJ. Bio-
syngas production from agro-industrial biomass residues by steam gasification.
Waste Manag 2016;58:221–9.
[3] Chang AC, Chang HF, Lin FJ, Lin KH, Chen CH. Biomass gasification for hydrogen
production. Int J Hydrogen Energy 2011;36:14252–60.
[4] Nipattummakul N, Ahmed II, Gupta AK, Kerdsuwan S. Hydrogen and syngas yield
from residual branches of oil palm tree using steam gasification. Int J Hydrogen
Energy 2011;36:3835–43.
[5] Cao JP, Huang X, Zhao XY, Wang BS, Meesuk S, Sato K, et al. Low-temperature
catalytic gasification of sewage sludge-derived volatiles to produce clean H2-rich
syngas over a nickel loaded on lignite char. Int J Hydrogen Energy 2014;39:9193–9.
[6] Dabai F, Paterson N, Millan M, Fennell P, Kandiyoti R. Tar formation and de-
struction in a fixed-bed reactor simulating downdraft gasification: equipment de-
velopment and characterization of tar-cracking products. Energy Fuels
2010;24:4560–70.
[7] Milne TA, Evans RJ, Abatzoglou N. Biomass gasification “tar”: their nature,
817
Fuel 216 (2018) 810–817
formation and conversion NREL/TP-570-25367 1998.
[8] Simell P, Kurkela E, Ståhlberg P, Hepola J. Catalytic hot gas cleaning of gasification
gas. Catal Today 1996;27:55–62.
[9] Świerczyński D, Courson C, Bedel L, Kiennemann A, Guille J. Characterization of
Ni–Fe/MgO/Olivine catalyst for fluidized bed steam gasification of biomass. Chem
Mater 2016;18:4025–32.
[10] Wang D, Yuan W, Ji W. Char and char-supported nickel catalysts for secondary
syngas cleanup and conditioning. Appl Energy 2011;88:1656–63.
[11] de Andrés JM, Narros A, Rodríguez ME. Air-steam gasification of sewage sludge in a
bubbling bed reactor: effect of alumina as a primary catalyst. Fuel Process Technol
2011;92:433–40.
[12] de Andrés JM, Roche E, Narros A, Rodríguez ME. Characterisation of tar from
sewage sludge gasification. Influence of gasifying conditions: temperature,
throughput, steam and use of primary catalysts. Fuel 2016;80:116–26.
[13] Pfeifer C, Rauch R, Hofbauer H. In-bed catalytic tar reduction in a dual fluidized
bed biomass steam gasifier. Ind Eng Chem Res 2004:1634–40.
[14] Garcia L, Salvador ML, Arauzo J, Bilbao R. Catalytic steam gasification of pine
sawdust. Effect of catalyst weight/biomass flow rate and steam/biomass ratios on
gas production and composition. Energy Fuels 1999;3:851–9.
[15] Virginie M, Adánez J, Courson C, De Diego LF, García-Labiano F, Niznansky D, et al.
Effect of Fe–olivine on the tar content during biomass gasification in a dual flui-
dized bed. Appl Catal B 2012;21:214–22.
[16] Ko TH, Chu H, Chaung LK. The sorption of hydrogen sulfide from hot syngas by
metal oxides over supports. Chemosphere 2005;8:467–74.
[17] Dayton D. A review of the literature on catalytic biomass tar destruction. National
Renewable Energy Laboratory (NREL); 1998.
[18] Devi L, Ptasinski KJ, Janssen FJJG. A review of the primary measures for tar
elimination in biomass gasification processes. Biomass Bioenergy 2003;24:125–40.
[19] Raman P, Ram NK, Gupta R. A dual fired downdraft gasifier system to produce
cleaner gas for power generation: design, development and performance analysis.
Energy 2013;54:302–14.
[20] Henriksen U, Ahrenfeldt J, Jensen TK, Gøbel B, Bentzen JD, Hindsgaul C, et al. The
design, construction and operation of a 75 kW two-stage gasifier. Energy
2006;1:1542–53.
[21] Mun TY, Kim JW, Kim JS. Air gasification of dried sewage sludge in a two-stage
gasifier: Part 1. The effects and reusability of additives on the removal of tar and
hydrogen production. Int J Hydrogen Energy 2013;8:5226–34.
[22] Mun TY, Kim JS. Air gasification of dried sewage sludge in a two-stage gasifier. Part
2: Calcined dolomite as a bed material and effect of moisture content of dried
sewage sludge for the hydrogen production and tar removal. Int J Hydrogen Energy
2013;38:5235–42.
[23] Mun TY, Cho MH, Kim JS. Air gasification of dried sewage sludge in a two-stage
gasifier. Part 3: Application of olivine as a bed material and nickel coated dis-
tributor for the production of a clean hydrogen-rich producer gas. Int J Hydrogen
Energy 2014;39:5634–43.
[24] Choi YK, Cho MH, Kim JS. Air gasification of dried sewage sludge in a two-stage
gasifier. Part 4: Application of additives including Ni-impregnated activated carbon
for the production of a tar-free and H2-rich producer gas with a low NH3 content. Int
J Hydrogen Energy 2016;38:5235–42.
[25] Kumar A, Jones DD, Hanna MA. Thermochemical biomass gasification: a review of
the current status of the technology. Energies 2009;2:556–81.
[26] Choi YK, Mun TY, Cho MH, Kim JS. Gasification of dried sewage sludge in a newly
developed three-stage gasifier: effect of each reactor temperature on the producer
gas composition and impurity removal. Energy 2016;14:121–8.
[27] Choi YK, Ko JH, Kim JS. A new type three-stage gasification of dried sewage sludge:
effects of equivalence ratio, weight ratio of activated carbon to feed, and feed rate
on gas composition and tar, NH3, and H2S removal and results of approximately 5 h
gasification. Energy 2017;18:139–46.
[28] Simell P, Stahlberg P, Kurkela E, Albrecht J, Deutsch S, Sjostrom K. Provisional
protocol for the sampling and analysis of tar and particulates in the gas from large-
scale biomass gasifier. Version 1998. Biomass Bioenergy 2000;18:19–38.
[29] Fukuyama H, Terai S, Uchida M, Cano JL, Ancheyta J. Active carbon catalyst for
heavy oil upgrading. Catal Today 2004;98(1-2 SPEC. ISS.):207–15.
[30] Gil J, Corella J, Aznar MP, Caballero MA. Biomass gasification in atmospheric and
bubbling fluidized bed: effect of the type of gasifying agent on the product dis-
tribution. Biomass Bioenergy 1999;17:389–403.
[31] Pakhare D, Spivey J. A review of dry (CO2) reforming of methane over noble metal
catalysts. Chem Soc Rev 2014;3:7813–37.
[32] Xu C, Donald J, Byambajav E, Ohtsuka Y. Recent advances in catalysts for hot-gas
removal of tar and NH3 from biomass Gasification. Fuel 2010;89:1784–95.
[33] Ko TH, Chu H, Chaung L. The sorption of hydrogen sulfide from hot syngas by metal
oxides over supports. Chemosphere 2005;58:467–74.
[34] Cheekatamarla PK, Lane A. Catalytic autothermal reforming of diesel fuel for hy-
drogen generation in fuel cells: II. Catalyst poisoning and characterization studies. J
Power Sources 2006;154:223–31.
[35] Wang Z, Flytzani-Stephanopoulos M. Cerium oxide-based sorbents for regenerative
hot reformate gas desulfurization. Energy Fuels 2005;19:2089–97.
[36] Yung MM, Jablonski WS, Magrini-Bair KA. Review of catalytic conditioning of
biomass-derived syngas. Energy Fuels 2009;3:1874–87.
[37] Di Gregorio F, Parrillo F, Salzano E, Cammarota F, Arena U. Removal of naphtha-
lene by activated carbons from hot gas. Chem Eng J 2016;291:244–53.
[38] Bandosz TJ. Effect of pore structure and surface chemistry of virgin activated car-
bons on removal of hydrogen sulfide. Carbon 1999;7:483–91.