Journal of Chemical Technology and Biotechnology J Chem Technol Biotechnol 79:570–577 (online: 2004)

DOI: 10.1002/jctb.1005

Control of H2S waste gas emissions with a

biological activated carbon filter
Ying-Chien Chung,1 Yu-Yen Lin2 and Ching-Ping Tseng2∗
1 Department of Industrial Engineering and Management, China Institute of Technology, Taipei, Taiwan 115, People’s Republic of China
2 Department of Biological Science and Technology, National Chiao Tung University, Hsinchu, Taiwan 300, People’s Republic of China

Abstract: The removal of high concentrations of H2 S from waste gases containing mixtures of H2 S and NH3
was studied using the pilot-scale biofilter. Granular activated carbon (GAC), selected as support material
in this study, demonstrated its high adsorption capacity for H2 S and good gas distribution. Extensive
tests to determine removal characteristics, removal efficiency, and removal capacity of high H2 S levels
and coexisting NH3 in the system were performed. In seeking the appropriate operating conditions,
the response surface methodology (RSM) was employed. H2 S removal capacities were evaluated by the
inoculated bacteria (biological conversion) and BDST (Bed Depth Service Time) methods (physical
adsorption). An average 98% removal efficiency for 0.083– 0.167 mg dm−3 of H2 S and 0.004– 0.021 mg dm−3
of NH3 gases was achieved during the operational period because of rapid physical adsorption by GAC
and subsequently an effective biological regeneration of GAC by inoculated Pseudomonas putida CH11
and Arthrobacter oxydans CH8. The results showed that H2 S removal efficiency for the system was not
affected by inlet NH3 concentrations. In addition, no acidification was observed in the BAC biofilter. High
buffer capacity and low moisture demand were also advantages of this system. The maximal inlet loading
and critical loading for the system were 18.9 and 7.7 g-H2 S m−3 h−1 , respectively. The results of this study
could be used as a guide for the further design and operation of industrial-scale systems.
 2004 Society of Chemical Industry

Keywords: biofilter; hydrogen sulfide; ammonia; granular activated carbon; waste gases

1 INTRODUCTION of H2 S in the coexistence of NH3 gas in a long-term

Hydrogen sulfide (H2 S) is a highly toxic chemical treatment.
while ammonia (NH3 ) can produce strong odors and Physical and chemical methods that have been
create visibility problems. H2 S and NH3 often coexist developed to purify gas pollutants from waste gases
and are released in large quantities from agricultural and wastewater include activated carbon adsorption,
activities and industrial processes such as livestock ozone oxidation, and incineration.12 – 14 Considering
farming, food and rubber processing, leather manu- treatment concentrations, physiochemical properties
facturing, wastewater treatment, landfills for waste dis- and cost of treating the pollutant, biological treatment
posal, and hog manure.1 – 4 Typical concentrations of is the best available control technology (BACT) for
NH3 and H2 S emitted from these activities and indus- odor treatment.15 The kinds of microbes and carriers
tries, including compost plants, livestock farming and present are crucial considerations in the development
fish processing, range from 0.004 to 0.042 mg dm−3 of waste gas treatment.15 Although appropriate
and from 0.007 to 0.138 mg dm−3 , respectively.1,5,6 and promising species, including Thiobacillus sp,
However, high concentrations of H2 S and low con- Xanthomonas sp and Pseudomonas sp have been
centrations of NH3 are often simultaneously emitted applied to H2 S removal,10,16,17 our previous studies
from rubber processing, leather manufacturing, and have shown that Pseudomonas putida CH11 has great
some wastewater treatment.7 In these mixed gases, potential to remove H2 S gas effectively because of
control of H2 S emission is difficult, yet important, its fast oxidation rate, low acid production, and high
because of its high toxicity and low threshold expo- removal capacity.7,10 Widely used carriers such as
sure limits.8 Some past studies have focused on H2 S soil, compost, peat, activated carbon, or mixtures of
removal alone and have achieved a high removal different materials have been examined. The results
efficiency,9 – 11 but few studies present removal char- indicate that soils have limited effectiveness since
acteristics and treatment data for high concentrations they are prone to short-circuiting and clogging,18

∗ Correspondence to: Ching-Ping Tseng, Department of Biological Science and Technology, National Chiao Tung University, Hsinchu,
Taiwan 300, People’s Republic of China
E-mail: [email protected]
Contract/grant sponsor: National Science Council
(Received 23 July 2003; revised version received 17 November 2003; accepted 6 December 2003)

 2004 Society of Chemical Industry. J Chem Technol Biotechnol 0268–2575/2004/$30.00 570

 10974660, 2004, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jctb.1005 by University Of Hertfordshire, Wiley Online Library on [08/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Control of H2 S waste gas emissions

and compost suffers from aging effects that decrease 0.48 g cm−3 , 1250 BET-m2 g−1 and 9, respectively.
the effectiveness of the biological system.19 While P putida CH11 or A oxydans CH8 were each grown
peat is a good carrier, it may produce pressure in 1000 cm3 nutrient broth for 1 day and then
loss and water control problems in a long-term harvested by centrifugation (8000 × g for 10 min).
treatment.20 Granular activated carbon (GAC), which The precipitate was put into a 10 dm3 PVC tank
is inoculated with mixed cells as a packing material, containing 7 dm3 nutrient broth. Prior to initiating
has become popular21,22 because it provides a the biofilter experiments, 2.7 kg of the granular
more uniform surface area and good resistance activated carbon (GAC) were separately mixed with
to crushing, allowing better operational control in the above solution for microbial attachment. During
areas such as gas distribution, pressure drop and the cultivation period, nutrient broth was added
pH.23 In addition, GAC provides higher pollutant every 3 days until the cell numbers of P putida
adsorption capacity than other inert carriers.22 Since CH11 and A oxydans CH8 reached 8.0 × 1010
biological oxidation occurs immediately after physical and 6.0 × 1010 cfu(g dry GAC)−1 , respectively. After
adsorption of activated carbon, minimizing the 28 days the GAC in the tank was then transferred into
treatment cost, the biological regeneration of GAC the biofilter.
gives the utilization of activated carbon a new
direction. 2.3 Apparatus and gas removal for continuous
In previous studies, we found that Pseudomonas operation
putida CH11 is effective in eliminating H2 S and that To investigate the capacity of GAC to adsorb
Arthrobacter oxydans CH8 is effective in removing H2 S, three glass columns (12 cm φ × 3 cm working
NH3 .10,24 In an effort to improve H2 S removal length) connected in series were packed with GAC
efficiency (90%) from a mixture with high levels of without microbes to perform the BDST (Bed Depth
H2 S and NH3 using Ca-alginate as a carrier,7 these Service Time) experiment. H2 S, 0.083 mg dm−3 , was
two species were simultaneously inoculated into the continuously supplied to these glass columns at
biofilter using activated carbon as a carrier. In this 500 dm3 h−1 . The desired concentration of H2 S at
study, the pilot-scale BAC biofilter was fed with the breakthrough was defined as 0.0033 mg dm−3 (ie:
various high H2 S concentrations and constant NH3 Ce /Co = 0.04).
loadings for a 172-day operating period. The removal A set-up of the pilot-scale experimental BAC
characteristics and appropriate operating conditions biofilter is shown in Fig 1. Two glass columns
for H2 S removal in the presence of NH3 were (12 cmφ × 40 cm working height) connected in series
investigated, and good control of H2 S and NH3 were packed with cell-laden GAC supported by a
emissions was achieved. perforated sieve plate at the bottom of the column to
allow the circulating liquid to flow out. The packed
volume and GAC dry weight in the biofilter were
2 MATERIALS AND METHODS 9.05 dm3 and 4.34 kg, respectively. Each glass column
2.1 Organism, cultivation and medium
preparation
The original pure-culture strains of heterotrophic
sulfur-oxidizer P putida CH11 and heterotrophic
ammonia-oxidizer A oxydans CH8 were isolated from
swine wastewater.10,24 Stock cultures were both grown
in nutrient broth at 30 ◦ C. The nutrient broth con-
tained yeast extracts (5 g dm−3 ), tryptone (10 g dm−3 )
and dextrose (2 g dm−3 ). In continuous-treatment
experiments, the inflow medium was supplied and
stored in the nutrient tank. The inflow medium con-
tained glucose (10 g dm−3 ), KH2 PO4 (4.08 g dm−3 ),
K2 HPO4 (5.22 g dm−3 ), NH4 Cl(0.4 g dm−3 ),
−3
MgCl2 ·6H2 O(0.2 g dm ) and Fe(III)-citrate
(0.01 g dm−3 )(C:N = 40:1). The final pH of the
medium was adjusted to neutral using 2 mol dm−3
NaOH or HCl. The buffer capacity in the inflow
medium was calculated as 0.033 (mol dm−3 pH−1 ).

2.2 Immobilization procedure

Granular activated carbon (extruded 4.5 mm particles,
from Cherng Tay Corporation Ltd in Taiwan) was
Figure 1. Schematic of the pilot scale BAC biofilter. 1, Glass column;
used as the support material for the biomass. The 2, flow meter; 3, NH3 gas cylinder; 4, H2 S gas cylinder; 5, air
main characteristics of the support material–bulking compressor; 6, nutrient tank; 7, pump; 8, regulator; 9, air filter;
density, specific surface area, and pH value–were 10, four-way connector.

J Chem Technol Biotechnol 79:570–577 (online: 2004) 571

 10974660, 2004, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jctb.1005 by University Of Hertfordshire, Wiley Online Library on [08/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Y-C Chung, Y-Y Lin, C-P Tseng

with two gas sampling ports, which spaced 20 cm Circulating liquid (20 cm3 ) was withdrawn from the
apart, were drilled along the column for measuring gas nutrient tank at regular intervals to measure the tank’s
concentrations. The flow meters and valves were used pH value. To determine the moisture content in the
for monitoring and controlling the gas flow through the GAC, about 5 g GAC were withdrawn, weighed, and
reactor. The pure H2 S and NH3 gases, supplied from dried over a 24-h period at 103 ± 0.5 ◦ C, and the
separate gas cylinders with flow rates ranging from 1.4 measurement was conducted 20 min after spraying
to 10.8 dm3 h−1 , were first diluted with compressed the aqueous medium. For cell number estimation,
air, which passed an air filter (pore size 0.2 µm, LIDA 0.5 g GAC was separately taken from sampling ports
3000–06, made in USA) and then flowed downward at the different depths and mixed with 5 cm3 sterile
through the biofilter at the top. The 7 dm−3 of inflow water. These samples were vortexed for 3 min, and
medium (see Section 2.1) stored in the nutrient tank the cell numbers in the filter were enumerated by
was recirculated by a peristaltic pump at 10 dm3 min−1 traditional plate-counting methods. In this case, the
for 6 min, once every 4 h to maintain the moisture of LB (Luria–Bertani) medium for heterotrophic micro-
the biofilter and supply nutrient to the attached cells. bial cultivation, the Cetrimide selective medium for
The peristaltic pump was connected to a spray nozzle P putida CH11,26 and the Hagedorn & Holt selec-
(spray angle 115◦ ; connecting thread 3/8’’; Chang tive medium for A oxydans spp were used.27 The
Der Co, Taiwan) to uniformly spray the medium on N-cetyl-N,N,N-trimethylammonium bromide in the
the filter bed’s surface. Glucose of 1% was added Cetrimide selective medium was able to largely inhibit
every 2 weeks. In the 172-day treatment experiment, the growth of the accompanying microbial flora except
different H2 S concentrations (0.083–0.167 mg dm−3 ) for Pseudomonas sp26 and the cycloheximide in the
under constant inlet NH3 loading (1.37 g-N m−3 h−1 ) Hagedorn & Holt selective medium was a selective
were introduced to the BAC biofilter at various flow agent for Arthrobacter sp.27 The inoculated plates grew
rates (180–1080 dm3 h−1 ) to evaluate the system’s for 2 days in an incubator at 26 ◦ C. The 16S rDNA
performance. sequence was applied after the selective medium was
used to ensure they were P putida CH11 and A oxydans
2.4 Experimental design CH8.
Response surface methodology (RSM), described
first by Box et al,25 is an experimental strategy for
seeking the appropriate operating conditions for a
multivariable system. The combined effect of inlet 3 RESULTS AND DISCUSSION
concentration and gas flow rate on H2 S removal was 3.1 H2 S adsorption by GAC
investigated using this methodology. In the present GAC is a good carrier for bacterial growth, but it
work, a central composite design (CCD) for two is also effective for direct adsorption of H2 S. The
variables was used to study the response pattern. The physical adsorption capacity of GAC to remove H2 S
full-factorial design consisting of a two-factor-two- is shown in Fig 2. The adsorption process is described
level pattern with 11 design points (nine combinations by Eqn (1):
with three replications of the center points) was used. A
multiple regression analysis was performed to obtain


No 1 Co
the coefficients, and these equations were used to t= X− ln −1 (1)
predict the response using the STATISTCA program. Co V KCo Ce

2.5 Analytical methods where Co and Ce are the inlet concentration and
Inlet H2 S gas concentrations in the reactor were peri- the desired concentration of gas at breakthrough
odically measured by gas detector tubes (Kitagawa, (mg dm−3 ), respectively; V , t and X are hydraulic
Japan) in the range of 0.00138–0.207 mg dm−3 . Out- loading (cm h−1 ), service time (h−1 ) and bed depth
let concentrations were continuously measured using (cm); No and K are adsorption capacity (mg dm−3 )
a Single Point Monitor (MDA Scientific, USA) in and adsorption rate constant (mg−1 dm3 h−1 ).
the range of 7 × 10−5 –2.1 × 10−4 mg dm−3 or peri- Equation constants were obtained from the slope
odically measured by gas detector tubes (Kitagawa, and the intercept of Fig 2 by a regression method.
Japan) in the range of 0.00138–0.083 mg dm−3 . Inlet The values of slope and intercept were 13.72 and
NH3 gas concentrations in the reactor were periodi- 0.069, respectively. The adsorption capacity (No ) and
cally measured by gas detector tubes (Kitagawa, Japan) the adsorption rate constant (K ) were 5057 mg dm−3
in the range of 0.0035–0.069 mg dm−3 . Outlet con- and 70.29 mg−1 dm3 h−1 , respectively. The adsorp-
centrations were continuously measured using a Single tion capacity expressed as dry weight of GAC
Point Monitor (MDA Scientific, USA) in the range was 11.78 mg-H2 S(g dry GAC)−1 . This indicates that
of 7 × 10−5 –7 × 10−4 mg dm−3 . To measure the pH, GAC has much greater adsorption capacity than Ca-
0.5 g GAC was withdrawn from the biofilter through alginate (6.8 × 10−4 mg-H2 S(g dry bead)−1 ).28 The
the appropriate sampling port and mixed with 5 cm3 data of H2 S adsorption capacity by pure GAC pro-
of distilled water. The sample was vortexed for 3 min, vide a reference value for the removal capacity of
and the pH was then determined using a pH meter. cell-laden GAC.

572 J Chem Technol Biotechnol 79:570–577 (online: 2004)

 10974660, 2004, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jctb.1005 by University Of Hertfordshire, Wiley Online Library on [08/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Control of H2 S waste gas emissions

1.2 H2 S loading returned to its original level, 80% removal

A. efficiency (RE) was achieved within 2 days, and 98%
1
removal efficiency was achieved in 5 days. Besides,
0.8 at the load used in this experiment, RE > 95% were
obtained at EBRT > 45 s and the RE dropped to
Ce/Co

0.6 about 60% at an EBRT of 30 s. During 6 months of

0.4 study, an average of 98% H2 S removal was demon-
column 1
column 2
strated. In comparison to other systems (compost
0.2 filter, three-phase fluidized bed bioreactor, or Ca-
column 3
0 alginate biofilter) under similar operating conditions,
0 20 40 60 80 100 120 140 160 180 200 the BAC biofilter exhibited much higher efficiency
Service time (h) than other biofilter systems,9,10,29 though it was infe-
160
rior to the parallel dual column filters packed with
B. compost.4 Additionally, the BAC biofilter also exhib-
140
Breakthrough time (hour)

ited a high buffer capacity for perturbed H2 S inlet.

120 Fast physical adsorption and subsequently effective
100 biological oxidation resulted in successful H2 S and
NH3 removal. Simultaneous bioregeneration of the
80
activated carbon was an additional economic advan-
60 tage to these systems. To understand the microbial
40 activity and distribution in the bioreactor, changes in
20 the cell numbers were examined. During the 172-day
experiment, variations in cell number ranged from
0
0 2 4 6 8 10 12 109 and 1010 cfu(g dry GAC)−1 at the different sam-
Column depth (cm) pling ports (data not shown). The profile of the cell
numbers did not show gradient change along the
Figure 2. Adsorption capacity of GAC to H2 S evaluated by three
column length. Hence, H2 S and NH3 gases accu-
identical GAC columns connected in series. (A) breakthrough curve
by BDST method and (B) linear correlation between packing height
mulated uniformly in the GAC or were distributed
and breakthrough time according to the equation using over the bioreactor and were subsequently degraded
breakthrough data. by the inoculated microbial population within the
matrix of carbon particles. The data of Table 1
3.2 Removal of high H2 S concentrations from show the cell numbers and distribution ratios of
mixed waste gases in continuous operation the inoculated cells (P putida and A oxydans) and
Because mixtures of H2 S at high concentrations and other heterotrophic bacteria in the middle zone of
NH3 at low concentrations are often simultaneously the BAC biofilter. The results indicated that sulfur-
emitted from various industries or manufacturing oxidizer P putida was still a dominant microorganism
processes, it is necessary to examine the removal effi- in the bacterial population and accounted for 92%,
ciencies or characteristics of a long-term treatment.7 94% and 92% of bacteria on the 60th, 120th, and
The H2 S removal efficiency profile in the BAC biofilter 172nd days, respectively. The ammonia-oxidizer A
during 172 days of operation is indicated in Fig 3(A). oxydans was second and accounted for 8%, 6%, and
When the system operated to pseudo steady state (ie 8%. The other heterotrophic bacteria were relatively
the rate of physical adsorption equals that of bio- few.
logical oxidation) on about the 22nd day, different Although the P putida and A oxydans cell num-
inlet H2 S loadings under constant inlet NH3 loading bers were almost equal at the beginning, the gas
(1.37 g-N m−3 h−1 ) were introduced into the biore- mixtures with a high concentration of H2 S and
actor to test the H2 S removal performance of the low concentration of NH3 would be the main rea-
system. Throughout all the operating periods, 100% son for the change in the microbial population. In
NH3 removal was achieved. The complete break- this system, we found that P putida maintained a
through time of H2 S concentration in this system high distribution ratio and cell number in the bac-
was about 44 days as estimated by the adsorption terial community during the process of intermittent
equation. Hence, physical adsorption of GAC was shock loading, and it brought a stable and high
theoretically responsible for 100% H2 S removal in removal efficiency. In addition, low concentrations of
the first 44 days. After that, the inoculated bacteria NH3 , in the range of 0.004–0.021 mg dm−3 , could
apparently contributed to complete removal (100%) be also effectively removed by inoculated A oxy-
from the 44th to the 59th day. After the 119th day, dans. To establish operating criteria for scale-up BAC
intermittent shock loadings were conducted to test biofilter, the relationship between inlet loading and
the response of the system. By day 119, the H2 S removal capacity for H2 S gas is indicated in Fig 3(B).
loading had been tripled (from 6.64 g-H2 S m−3 h−1 As shown in Fig 3(B), the relationship curve first
to 19.92 g-H2 S m−3 h−1 ) for 4 days. Although removal rose and then leveled off to a maximum level. The
efficiency (>98%) was down to 55%, once the inlet critical loading (ie complete removal capacity) was

J Chem Technol Biotechnol 79:570–577 (online: 2004) 573

 10974660, 2004, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jctb.1005 by University Of Hertfordshire, Wiley Online Library on [08/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Y-C Chung, Y-Y Lin, C-P Tseng

40
A.
100 36

Inlet H2S loading (g-H2S m-3 h-1)

32

Removal efficiency (%)

80
28
24
60
20
Hydrogen sulfide 16
40 Ammonia
Inlet loading 12

20 8
4
0 0
0 20 40 60 80 100 120 140 160 180
Operation time (d)

20
B.
H2S removal capacity (g-H2S m-3 h-1)

18
16 100% removal curve
14
12
10
8
6 Maximum inlet loading
4
2 Crictical loading

0
0 2 4 6 8 10 12 14 16 18 20
Inlet H2S loading (g-H2S m-3 h-1)

Figure 3. (A) H2 S, NH3 removal efficiency in the BAC biofilter under constant inlet NH3 loading (1.37 g-N m−3 h−1 ) during 172-day operation at
room temperature and (B) relationship between inlet loading and removal capacity for H2 S gas.

Table 1. Cell numbers and distribution ratios of inoculated cells and other heterotrophic bacteria in the middle zone of BAC biofilter

Day

Straina 0 60 120 172

P putida 8.0 × 1010 (57%) 7.4 × 109 (92%) 5.7 × 109 (94%) 6.3 × 109 (92%)
A oxydans 6.0 × 1010 (43%) 6.3 × 108 (8%) 3.8 × 108 (6%) 5.4 × 108 (8%)
Other 0 (0%)b 7.4 × 105b 3.6 × 105b 5.2 × 105b
a Indicates the unit as cfu (g dry GAC)−1 .
b indicates the distribution ratio’s lower than 0.01%.

determined as 7.7 g-H2 S m−3 h−1 . The extrapolated response to temperature variation, the H2 S removal
correlation line suggested a maximum inlet load- efficiencies with temperature changes at gas reten-
ing of 18.9 g-H2 S m−3 h−1 (0.85 g-H2 S kg−1 day−1 ). tion time of 45 s are shown in Fig 4. The experiment
Compared with biological systems with peat support, was conducted on the same biofilter after a 172-
values of which were 0.102, 0.403, and 0.468 g- day operating period. H2 S loading was controlled
H2 S kg−1 day−1 ,30 – 32 the maximum inlet loading of at 9.64 g-H2 S m−3 h−1 , and NH3 loadings were in
the BAC biofilter was better. Thus, efficient and the range of 0.389–2.04 g-NH3 m−3 h−1 . Since the
acceptable H2 S removal can be accomplished through optimum temperature for P putida CH11 growth is
adjusting the flow rate and inlet concentration based 26 ◦ C,10 the removal efficiency slightly decreased from
on the correlation presented in Fig 3(B). 99% to 96% when temperature rose from 26 ◦ C to
45 ◦ C. Surprisingly, 96% removal of H2 S was still
3.3 Effect of temperature on H2 S removal maintained even with the temperature controlled at
Temperature is an important factor that will 45 ◦ C for another 6 days. When the temperature of the
affect physical absorption, biological oxidation, and system was returned to 26 ◦ C, the removal efficiency
especially microbial growth. To estimate the system was 99% within 4 days. These results suggest that

574 J Chem Technol Biotechnol 79:570–577 (online: 2004)

 10974660, 2004, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jctb.1005 by University Of Hertfordshire, Wiley Online Library on [08/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Control of H2 S waste gas emissions

a temporary biological deactivation did not result in 9

complete loss of removal capacity of the BAC biofilter
because the adsorption process might remove H2 S gas. 8
Therefore, the BAC biofilter possessed high adaptabil-
ity to temperature shock, and this property would favor
7

pH value
its application for H2 S removal in the field.

6
3.4 pH and moisture change in the BAC biofilter
pH in leachate
Acidification has often been an obstacle to traditional Sampler 1
biofilter methods for acid gas treatment.4 Hence, pH 5 Sampler 2
Sampler 3
control at a constant level in the bioreactor and Sampler 4
leachate was very important for biofilter function. 4
During the 172-day treatment in this study, the pH 0 20 40 60 80 100 120 140 160 180
values in the filter bed and leachate were determined Time (d)
every 4 days, and the results are shown in Fig 5. Figure 5. Change of pH values in the leachate and the filter operated
The increasing acid gas H2 S at constant inlet NH3 with BAC biofilter for 172 days.
loading resulted in a decrease of pH in the leachate.
The lowest pH value of 4.5 was observed in the moisture content in GAC was required to efficiently
leachate on the 120th day because the shock loading remove waste gas. The moisture content of GAC
of inlet H2 S was conducted on the 120th day, and at the different filter depths was determined every
then 0.1 mol dm−3 NaOH solution was added into 5 days, and the results are shown in Fig 6. Moisture
the nutrient tank to restore the pH to neutral. The content varied from 34% to 42% with the average at
acidification did not occur in the filter because the 38%. By comparison, the transitional biofilter required
changes in pH value were between 6.0 and 8.5. Since a 40–60% moisture content to maintain biological
the pH of the activated carbon was 9, it was able activity.34 The BAC biofilter is energy-saving and
to stabilize the pH of the filter bed and provide the highly efficient.
system with high alkalinity following the addition of
acidic H2 S gas. The presence of ammonia in the waste
3.5 Effect of NH3 on H2 S removal
gas provided another neutralizing effect. Therefore,
To understand the effect of NH3 coexistence on H2 S
the profile of the pH showed a gradient change in
removal from mixed waste gases in the BAC biofilter,
the axial direction of the filter bed. These results
different NH3 concentrations (0.007–0.084 mg dm−3 )
were due to accumulation of more acid products on
were introduced at 360 dm3 h−1 with 0.138 mg dm−3
the lower layer (data not shown). This phenomenon
H2 S. The study was a separate experiment and
was attributed to the combined effect of liquid flow
was conducted after a 172-day operating period.
direction and gravity.
Compared with the previous studies, the H2 S removal
Microbial activity and mass-transfer rate often
efficiency gradually fell to 88% in the Ca-alginate
depend on the moisture content in or around the
immobilization system because of slight alkalization
support material. High moisture could lead to a serious
from NH3 coexistence.7 By contrast, H2 S removal
problem, such as the formation of stagnant zones with
by the BAC biofilter was not influenced by NH3
diffusion limitations and possible anaerobic conditions
concentrations under these operating conditions. A
or increased pressure drop.1,33 Thus, an appropriate
removal efficiency of 99% was achieved in all operating

50

100
45
Removal efficiency (%)

Moisture content (%)

95 40
26 °C 35 °C 40 °C 45 °C 26 °C
90 35

30
85 Hydrogen sulfide Sampler 1
Ammonia Sampler 2
25 Sampler 3
80 Sampler 4
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
20
Time (d) 0 20 40 60 80 100 120 140 160 180
Time (d)
Figure 4. Effect of temperature on H2 S and NH3 removal efficiency. , average moisture content (38%)
H2 S concentration was controlled at 0.138 mg dm−3 . NH3
concentrations were in the range of 0.004–0.021 mg dm−3 , and 100% Figure 6. Change of moisture content at different sampling ports in
NH3 removal efficiency was achieved. the BAC biofilter.

J Chem Technol Biotechnol 79:570–577 (online: 2004) 575

 10974660, 2004, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jctb.1005 by University Of Hertfordshire, Wiley Online Library on [08/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Y-C Chung, Y-Y Lin, C-P Tseng

conditions. Thus, the results of this study suggest 4 CONCLUSIONS

that using activated carbon as a packing material or Control of high concentrations of H2 S together with
a temporary gas storage tank would facilitate high low concentrations of NH3 emitted from industrial
removal efficiencies. waste gas treatments is an important process because
H2 S has high toxicity and a low threshold limit. In
3.6 Comprehensive evaluation of concentration this study, we demonstrate that the BAC biofilter
and flow rate for H2 S removal can effectively remove a high concentration of H2 S
Because the most important physical factors affecting gas from mixed waste gases containing H2 S and
H2 S removal are inlet H2 S concentrations and gas NH3 . The system achieved an average 98% removal
flow rates, experiments were performed at different efficiency of H2 S during a 172-day operating period
inlet concentrations and flow rate combinations, when NH3 concentrations were between 0.004 and
and suitable combinations for emission limits were 0.021 mg dm−3 . Additionally, the system shows the
determined when NH3 concentration fluctuated potential for high H2 S removal efficiency even
within 0.004–0.021 mg dm−3 . Using the results of with temperatures as high as 45 ◦ C. No significant
the experiments, the following equation giving the acidification phenomenon occurred in this system
relative H2 S removal efficiency as the function of inlet during H2 S treatment. The low moisture demand
concentration and flow rate was obtained: of the BAC biofilter was another advantage. A set of
the operating combinations was further established for
application in the field. Therefore, the results of this
y = 101.3621 − 0.00102x1 − 0.0054x2 study suggested that the BAC biofilter has significant
− 0.000016x1 x2 (2) potential to treat H2 S gas from mixed waste gases
containing NH3 .

where x1 = H2 S concentration and x2 = gas flow rate.

According to the above equation, all the factors ACKNOWLEDGEMENT
have negative effects. The equation also shows that x2 The work was partially supported by Grant NSC-90-
(gas flow rate) is the most significant factor, with 2317-B-009-001 from the National Science Council.
its coefficient effect being most pronounced. The
contour plot of Fig 7 explains the behavior of the
system. If 100% removal efficiency was achieved REFERENCES
at 0.07 mg dm−3 of inlet H2 S, the gas flow rate 1 Langenhove HV, Wuyts E and Schamp N, Elimination of
should be controlled below 225 dm3 h−1 . In addition,
hydrogen sulphide from odorous air by a wood bark biofilter.
Water Res 20:1471–1476 (1986).
99% removal was obtained when the gas flow rate 2 Eikum AS and Storhang R, Odour problems related to waste
was controlled below 375 dm3 h−1 . By the developed water and sludge treatment in. Odour Prevention and Control of
equation or contour plot, a set of suitable operating Organic Sludge and Livestock Farming, ed by Neilsen VC,
combinations was determined and could be further Voorburg JH and Hermite PL. Elsevier Applied Science
Publishers, London, pp 12–18 (1986).
applied in the field with gases containing high 3 Devinny JS, Deshusses MA and Webster TS, Biofiltration for
concentrations of H2 S and coexistent NH3 . Air Pollutant Control. Lewis Publishers, New York, pp 7–8
(1999).
4 Yang Y and Allen ER, Biofiltration control of hydrogen sulphide
H2S Removal efficiency (%) 1. Design and operational parameters. J Air Waste Manage
44:863–868 (1994).
700 96 95 94 93 5 Cho KS, Hirai M and Shoda M, Enhanced removal efficiency
of malodorous gases in a pilot-scale peat biofilter inoculated
94 with Thiobacillus thioparus DW44. J Ferment Bioeng 73:46–50
96 95
97 (1992).
600
6 Chung YC, Huang C and Tseng CP, Reduction of H2 S/NH3
96 95
Flow rate (dm3 h-1)

97 production from pig feces by controlling environmental

98 conditions. J Environ Sci Health Part A 31:139–155 (1996).
500 97 96 7 Chung YC, Huang C, Tseng CP and Pan JR, Biotreatment of
98 H2 S- and NH3 - containing waste gases by co-immobilized
97 cells biofilter. Chemosphere 41:329–336 (2000).
400 98 8 Buchman RE and Gibbons NE, Bergey’s Manual of Determina-
99
99 98 tive Bacteriology. Williams and Wilkins Co, Baltimore, MD,
pp 52–54 (1974).
99
300 9 Rands MB, Cooper DE, Woo CP, Fletcher GC and Rolfe KA,
99 Compost filters for H2 S removal from anaerobic digestion and
100 rendering exhausts. J Water Pollut Control Feder 53:185–189
100
200 100 (1981).
100
10 Chung YC, Huang C and Tseng CP, Biodegradation of hydro-
0.06 0.08 0.10 0.12 0.14 0.16
gen sulphide by a laboratory-scale immobilized Pseudomonas
Inlet H2S concentration (mg dm-3) putida CH11 biofilter. Biotechnol Prog 12:773–778 (1996).
11 Oh KJ, Kim D and Lee IH, Development of effective hydrogen
Figure 7. Contour plot showing H2 S removal efficiency in response sulphide removing equipment using Thiobacillus sp IW.
to varying concentration and flow rate. Environ Pollut 99:87–92 (1998).

576 J Chem Technol Biotechnol 79:570–577 (online: 2004)

 10974660, 2004, 6, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jctb.1005 by University Of Hertfordshire, Wiley Online Library on [08/11/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Control of H2 S waste gas emissions

12 Eby HJ and Wilson GB, Poultry house dust, odour and 22 Graham JR, GAC based gas phase biofiltration, in Proceedings of
their mechanical removal, in Agricultural Waste Management the 1996 Conference on Biofiltration, ed by Reynolds FE. The
Proceedings, Cornell University Conference on Agricultural Waste Reynolds Group, Tustin, CA, pp 85–90 (1996).
Management, Syracuse, NY, pp 303–308 (1969). 23 Medina VF, Webster T and Devinny JS, Treatment of gasoline
13 Barth CL, Elliott FL and Melvin SW, Using odour control residuals by granular activated carbon based biological
technology to support animal agriculture. Trans ASAE filtration. J Environ Sci Health Part A 30:407–412 (1995).
27:859–864 (1984). 24 Chung YC, Huang C and Tseng CP, Biotreatment of ammonia
14 Mannebeck H, Covering manure storing tanks to control odour, from air by an immobilized Arthrobacter oxydans CH8 biofilter.
in Odour Prevention and Control of Organic Sludge and Livestock Biotechnol Prog 13:794–798 (1997).
Farming, ed by Neilsen VC, Voorburg JH and Hermite PL. 25 Box GEP, Hunter WG and Hunter JS, Statistics for Experiments.
Elsevier Applied Science Publishers, London, pp 188–192 John Wiley and Sons, New York, pp 374–418 (1978).
(1986). 26 Mossel DAA and Indacochea L, A new cetrimide medium for
15 Leson G and Winer AM, Biofiltration: an innovative air the detection of Pseudomonas aeruginosa. J Medical Microbiol
pollution control technology for VOC emission. J Air Waste 4:380–382 (1971).
Manage 41:1045–1054 (1991). 27 Hagedorn C and Holt JG, A nutritional and taxonomic survey of
16 Wainwright M, Sulfur oxidation in soil. Adv Agron 37:350–396 Arthrobacter soil isolates. Can J Microbiol 21:353–361 (1975).
(1984). 28 Chung YC, Huang C and Tseng CP, Removal of hydrogen
17 Cho KS, Hirai M and Shoda M, Degradation characteristics sulphide by immobilized Thiobacillus sp strain CH11 in a
of hydrogen sulphide, methanethiol, dimethyl sulphide biofilter. J Chem Technol Biotechnol 68:58–62 (1997).
and dimethyl disulphide by Thiobacillus thioparus DW44 29 Kim KR, Oh KJ, Park KY and Kim D, Removal of hydrogen
isolated from peat biofilter. J Ferment Bioeng 71:384–389 sulphide and methylmercaptan using Thiobacillus in a
(1991). three phase fluidized bed bioreactor. J Microbiol Biotechnol
18 Carlson DA and Leisner CP, Soil beds for the control of 9:265–270 (1999).
sewage odors. J Water Pollut Control Feder 38:829–834 30 Furusawa N, Togashi I, Hirai M, Shoda M and Kubota H,
(1966). Removal of hydrogen sulphide by a biofilter with fibrous
19 Langenhove HV, Bendinger B, Oberthur R and Schamp N, peat. J Ferment Technol 62:589–593 (1984).
Organic sulfur compounds: persistent odorants in biological 31 Wada A, Shoda M, Kubota H, Kobayashi T, Yoko KF and
treatment of complex waste gases, in Biotechniques for Air Kuraishi H, Characteristics of H2 S oxidizing bacteria inhab-
Pollution Abatement and Odour Control Policies, ed by Dragt AJ iting a peat biofilter. J Ferment Technol 64:161–168 (1986).
and Ham JV. Elsevier Science Publishers, Amsterdam, pp 32 Cho KS, Hirai M and Shoda M, Degradation of hydrogen
177–182 (1992). sulphide by Xanthomonas sp strain DY44 isolated from peat.
20 Tanji Y, Kanagawa T and Mikami E, Removal of dimethyl Appl Environ Microbiol 58:1183–1189 (1992).
sulphide, methyl mercaptan, and hydrogen sulphide by 33 Ottengraf SPP and Van den Oever AHC, Kinetics of organic
immobilized Thiobacillus thioparus TK-m. J Ferment Bioeng compound removal from waste gases with a biological filter.
67:280–285 (1989). Biotechnol Bioeng 25:3089–3102 (1983).
21 Weber FJ and Hartmans S, Use of activated carbon as a buffer 34 Sabo F, Development and testing of high-efficiency biofilters,
in biofiltration of waste gases with fluctuating concentration in Proceedings of the 86th Annual Meeting of the Air and Waste
of toluene. Appl Microbiol Biotechnol 43:365–372 (1995). Management Association, Denver, Colorado (1993).

J Chem Technol Biotechnol 79:570–577 (online: 2004) 577