Removal of hydrogen sulfide from gas streams

using biological processes - A review

M. Syed, G. Soreanu*, P. Falletta and M. Béland

Wastewater Technology Centre, Environment Canada, 867 Lakeshore Road, Burlington, Ontario L7R 4A6, Canada. *E-mail:
[email protected]

Syed, M., Soreanu, G., Falletta, P. and Béland M. 2006. Removal of rigoureux de l’apport d’oxygène pour assurer la production de soufre
hydrogen sulfide from gas streams using biological processes - A élémentaire au lieu de sulfate. Ces réacteurs gagnent une certaine
review. Canadian Biosystems Engineering/Le génie des biosystèmes popularité dans la purification de divers flux gazeux contenant du H2S
au Canada 48: 2.1 - 2.14. Hydrogen sulfide (H2S) is a toxic and comme le biogaz, les gaz d’échappement des usines de traitement des
odorous compound present in biogas produced by the anaerobic eaux usées, des fermes etc. Le garnissage du biofiltre est peu coûteux
digestion of biosolids and other organic materials. Due to its corrosive et peut contenir les nutriments nécessaires pour le développement et
nature in internal combustion engines, biological hydrogen sulfide soutien des populations microbiennes. Les besoins de recherche dans
removal processes are being investigated to overcome the chemical and ce domaine inclus l’optimisation de la biofiltration anaérobie, le
disposal costs associated with existing chemically-based removal développement d’un système qui prend avantage des caractéristiques
processes. Both phototrophic and chemotrophic bacteria are suitable des bactéries phototrophiques et chimiotrophiques et d’un système qui
candidate microorganisms for hydrogen sulfide bioxidation. pourrait réaliser l’élimination simultanée du sulfure d'hydrogène et du
Phototrophic C. limicola is an ideal bacterium in these biological siloxane. Mots-clés: sulfure d'hydrogène, phototrophique,
removal processes due to its ability to grow under anaerobic conditions chimiotrophique, oxydation biologique, biofiltration, épuration
using only inorganic substrates and a light source and its efficient biologique.
extracellular production of elemental sulfur from H2S. Phototrophic
fixed-film reactors are an interesting concept for cost-effective H2S INTRODUCTION
removal from biogas due to their ability to operate for long periods of
time without requiring a biomass separation step and their ability to Hydrogen sulfide is present in biogas produced during the
operate under higher and variable loadings. However, a light source is anaerobic digestion of biodegradable substances. It is produced
one of the key constraints for this process. Chemotrophic bacteria can from the degradation of proteins and other sulfur containing
also be used in fixed-film reactors to produce elemental sulfur instead compounds present in the organic feed stock to the digester. The
of sulfate under controlled oxygen conditions. These bioscrubbers are concentration of hydrogen sulfide in the biogas depends on the
gaining acceptance for treating hydrogen sulfide containing gases from feedstock and varies between 0.1 to 2% (Lastella et al. 2002;
a wide variety of sources such as biogas, off-gases from wastewater Boyd 2000). Generally, biogas produced from manure and
treatment plants, livestock farms, etc. The biofilter medium is
inexpensive and may contain sufficient micro-nutrients for the protein rich industrial wastes contain higher amounts of
microbial communities. Future research needs include the optimization hydrogen sulfide (Schieder et al. 2003). Considerable amounts
of the anaerobic biofiltration process, the development of a system of hydrogen sulfide are also emitted from industrial activities
combining the advantages of phototrophic and chemotrophic bacteria such as petroleum refining (Henshaw et al. 1999), pulp and
and the possible co-removal of siloxanes within this process. paper manufacturing (Wani et al. 1999), food processing,
Keywords: hydrogen sulfide, phototrophic, chemotrophic, biological livestock farming (Chung et al. 2001), and natural gas
oxidation, biofiltration, bioscrubbing. processing (Kim et al. 1992). It is also found in landfill biogas
Le sulfure d'hydrogène (H2S) est un composé toxique et odorant and is the principal odorous component in off-gases from
présent dans le biogaz produit par la digestion anaérobique des wastewater collection and treatment facilities (Cox and
biosolides et de autres matières organiques. Ce composé peut causer Deshusses 2001). Biogas derived from these waste stabilization
la corrosion précoce des moteurs à combustion interne qui brule le processes is not usually used as a renewable energy source, but
biogaz pour générer de l’énergie. Récemment, certains procédés
biologiques sont visés pour surmonter les coûts élevés relié aux rather flared off as excess gas when it is not used for space and
processus chimiques d'élimination du sulfure d'hydrogène (coûts des process heating (Tchobanoglous et al. 2003). One of the biggest
produits chimiques et de l’élimination des déchets). Les factors limiting the use of biogas is related to the hydrogen
microorganismes responsables pour la bio-oxydation du H2S peuvent sulfide it contains, which is very corrosive to internal
être de nature phototrophique ou chimiotrophique. Due à sa capacité combustion engines (Tchobanoglous et al. 2003; Ross et al.
de croissance rapide et de sa production extracellulaire du soufre 1996). Currently, most commercial technologies for the removal
élémentaire en utilisant seulement des substrats inorganiques en milieu of H2S are chemically based and expensive to operate (Monteith
anaérobique, la bactérie phototrophique C. limicola est suggérée. Les et al. 2005; Gabriel and Deshusses 2003; Cha et al. 1999)
réacteurs à biomasse phototrophique fixée sont intéressants pour la
thereby negating all of the financial incentives associated with
conversion du H2S grâce à leur facilité de séparer le soufre du biogaz
et de leur capacité élevée d’opérer sous des charges et des conditions potential revenues from energy produced in a cogeneration
variables. Cependant, l'alimentation en énergie lumineuse est l’une des plant.
contraintes principales pour l’avancement de ce procédé avec la In addition to its unpleasant odor, hydrogen sulfide (H2S)
bactérie C. limicola. Les bactéries chimiotrophiques peuvent aussi être gas is highly toxic (Roth 1993). Upon inhalation, hydrogen
utilisées dans les réacteurs à biomasse fixée si l’on pratique un contrôle sulfide reacts with enzymes in the bloodstream and inhibits

Volume 48 2006 CANADIAN BIOSYSTEMS ENGINEERING 2.1

 Addition of organic oxidation of H2S and S0 to form SO42-. In the deep
matter, sulfates anaerobic zone, anaerobic decomposition of organic
Chemoautotrophs, matter occurs and H2S is produced. In the upper
water surface, oxygen H2S + CO2 + O2 CH2O + SO 42- anaerobic zone where light can still penetrate and H2S
available, low sulfide conc. Sº + CO2 + O2 CH2O + SO4
2-
is available, growth of phototrophic bacteria occurs.
These bacteria find suitable conditions for growth only
Anaerobic, light penetrates H2S + CO2 Sº + CH2O in a narrow zone of overlap since sulfide and light
photoautotrophs occur in opposite gradients. In these narrow layers,
they obtain reducing electrons from either H2S or S0
(Overmann 2001).
Anaerobic, high sulfide H2S + CO2 CH2O + SO42-
Organic Sulfur
The desirable bacteria to be used in a bioprocess to
H2S + CO2
concentration
convert H2S to S0 should possess the following basic
features: reliable capability of converting H2S to S0,
minimum nutrient inputs, and easy separation of S0
Fig. 1. Cycling of sulfur in a sulfuretum (adapted from Anderson from the biomass. Relevant photoautotrophs and
1978). chemotrophs are discussed below.
cellular respiration resulting in pulmonary paralysis, sudden Photoautotrophs Studies on microbial ecology associated with
collapse, and death. Continuous exposure to low (15-50 ppm) phototrophic bacteria have shown that a species of green sulfur
concentrations will generally cause irritation to mucous bacteria (GSB) Cholorobium limicola (originally called
membranes and may also cause headaches, dizziness, and Cholorobium limicola forma thiosulfatophilum (Larsen 1952))
nausea. Higher concentrations (200-300 ppm) may result in is the most suitable for sulfide removal and satisfies the criteria
respiratory arrest leading to coma and unconsciousness. for a desirable bacterium (Syed and Henshaw 2003).
Exposures for more than 30 minutes at concentrations greater Cholorobium limicola is capable of oxidizing sulfide to
than 700 ppm have been fatal (MSDS 1996). elemental sulfur, requires only light, CO2, and inorganic
Scrubbing, carbon adsorption, and chemical and thermal nutrients for growth and is strictly anaerobic. GSB are non-
oxidation are used for H2S removal from gas fluxes motile and deposit elemental sulfur extracellularly. This feature
(Shareefdeen et al. 2002; Cha et al. 1999). The Claus process is makes GSB suitable where the recovery of elemental sulfur
one of the most popular processes used for hydrogen sulfide from sulfide-containing wastewater is desired. The overall
removal with sulfur recovery on an industrial scale (Kim and photochemical reaction by which GSB oxidizes S2- to S0 while
Chang 1991; Cooper and Alley 1986). In the first step of the reducing CO2 to carbohydrates is (van Niel 1931):
Claus process, H2S is partially oxidized to SO2 with air. The
H2S/SO2 mixture is then reacted over a bauxite catalyst to yield
light energy
(
2nH2 S + nCO2 → 2nS 0 + n CH2 O + nH2 O ) (1)
elemental sulfur (S0) and water. Normally 90-95% of the H2S is Studies involving phototrophic bacteria are summarized in
converted to S0. The remaining H2S is either oxidized to SO2 or Table 1. Cork et al. (1985) introduced the concept of the "van
converted to sulfur in an off-gas treatment unit (Sittig 1978). Niel curve" by plotting the reactor feed rate as a function of
The chemical H2S removal processes are expensive due to irradiance (W/m2) for their batch-fed reactor system (Fig. 2).
high chemical requirements, energy and disposal costs (Buisman The curve describes the relationship between S2- loading rate
et al. 1989). For these reasons, biological treatment methods for and light intensity (radiant flux). When light intensity and
hydrogen sulfide removal are desirable as an alternative to sulfide flow rate were adjusted to a point on the curve (balanced
chemical treatment (Sercu et al. 2005; Oyarzún et al. 2003). A loading), all of the sulfide introduced to the reactor was
biological hydrogen sulfide removal process has the potential to oxidized to elemental sulfur without the formation of sulfate.
overcome some or all of the disadvantages of chemical Under sulfide overloading conditions (to the right of the curve),
processes (Elias et al. 2002; Kim et al. 2002). light energy was not sufficient and sulfide accumulated in the
This paper reviews current biological hydrogen sulfide reactor. When the reactor was in a sulfide underloading
removal processes in practice and identifies research needs for condition (to the left of the curve), the surplus light caused the
potential process improvement. formation of sulfate as shown by Eq. 2 (Larsen 1952).

LITERATURE REVIEW nH2 S + 2nCO2 + 2nH2 O 

light energy
→

Bacteria used in bioreactors nSO42 − + 2nH + + 2n( CH 2 O) (2)

Figure 1 shows conversions of different species of sulfur by
Therefore, only when the bioreactor system is adjusted to
naturally occurring bacteria where a complete oxidation to
operate "on the curve", sulfide removal is complete and a
elemental sulfur is occurring. Such a situation often occurs in
maximum amount of elemental sulfur is produced.
nature and is called a sulfuretum. A typical example is a pond in
autumn where fallen leaves are the source of organic matter The in vivo light absorption spectrum of C. limicola exhibits
(Postgate 1968). Different bacteria tend to live in areas of the light absorption between 350 and 850 nm (Stanier et al. 1986)
pond where their particular capabilities provide them with an with a peak at 760 nm. The authors describe two different
ecological niche. Near the water surface, chemotrophic bacteria conditions under which the quality of light available is different.
dominate where they can obtain their energy from the aerobic In shallow ponds, relatively rich in organic matter, except near

2.2 LE GÉNIE DES BIOSYSTÈMES AU CANADA SYED et al.

Table 1. Research conducted in hydrogen sulfide removal using photoautotrophs (adapted from Syed 2003).

Volume# Influent S2- loading Removal Irradiance‡

Reference Configuration*
(L) [H2S] (mg h-1 L-1) efficiency† (%) (W/m2)

Kobayashi et al. (1983) FF, U 8 16 mg/L in liquid 0.59-1.27 81-92 NQ

Kobayashi et al. (1983) FF, plug 0.1 19-24 mg/L in liquid 102-125 100 NQ
Cork et al. (1985) SG, CSTR 0.8 Gas, concentration unknown 74-109 100 150-2000
Maka and Cork (1990) SG, CSTR 0.8 1-2 mM in gas 32-64 90-100 139
Kim et al. (1991) SG, CSTR 4 2.1 mM in gas 61 >99 1200
Kim et al. (1992) SG, CSTR 4 2.1 mM in gas 64 100 1750
Kim et al. (1996) SG, CSTR 11.9 1.45-1.87 mM in gas 14.6-19 99.8 15.2
Basu et al. (1996) SG, CSTR 1.25 25,000 ppm in gas 94.4 >96.6 ID
Henshaw et al (1997) SG, CSTR 13.7 90-550 mg/L in liquid 2.1-5.6 >90 258
Henshaw and Zhu (2001) FF 0.02 141-380 mg/L in liquid 111-286 82-100 25.4
Syed and Henshaw (2003) FF 0.0048 91-164 mg/L in liquid 1323-1451 100 152

* CSTR = continuously stirred tank reactor; FF = fixed-film; SG = suspended-growth; U = upflow

#
Volume = wet volume of reactor
†
Removal efficiency = (S=in - S=out)/S=in
‡
ID = insufficient data to calculate; NQ = not quantified

layer. In this case, the overlying water column acts as a light

2.0 filter, transmitting only green and blue-green light, of
wavelengths between 450 and 550 nm. Carotenoids become the
dominant light harvesting pigments and the GSB in this
environment typically contain a very high carotenoid content.
1.5
Chemotrophs A number of chemotrophs are suitable for the
Irradiance (KW/m2)

biodegradation of H2S. These bacteria grow and produce new

Region I
S 2- underloading cell material by using inorganic carbon (CO2) as a carbon source
1.0 and chemical energy from the oxidation of reduced inorganic
compounds such as H2S. In the presence of reduced organic
carbon sources (glucose, amino acids, etc.), some of these
Region II bacteria (so-called mixotrophic microorganisms) can grow
0.5 heterotrophically, using the organic carbon as a carbon source
S 2- overloading
and an inorganic compound as an energy source (Prescott et al.
2003).
0.0 Biodegradation of H2S by chemotrophs occurs in aerobic
0 1 2 3 4 conditions with O2 as an electron acceptor or in anaerobic
conditions with alternative electron acceptors (e.g. nitrate),
-1 -1
H2S utilized (mmol L h ) depending on the type of bacteria (Prescott et al. 2003).
Examples of energy sources for representative chemotrophs are
Fig. 2. van Niel Curve (adapted from Cork et al. 1985). presented in Table 2.
The metabolism of species such as Thiobacillus,
the air-water interface, the water is oxygen-free allowing green Thermothrix, Thiothrix, Beggiato has been intensively studied
sulfur bacteria to grow close to the water surface. There they for oxidation of inorganic (elemental sulfur, hydrogen sulfide,
obtain light of long wavelength, which is transmitted through the thiosulfate) or organic (methanethiol, dimethylsulfide,
overlying aerobic phototrophs, and the light, in the far red and dimethyldisulfide) sulfur compounds (Stanier et al. 1986).
near-infrared regions, used by the GSB for photosynthesis is These microorganisms grow in soil, aquatic habitats, activated
almost entirely absorbed by bacteriochlorophylls. The second sludge systems, etc. under aerobic, microaerophilic, and
environment occurs in lakes where a warmer, aerobic layer anaerobic conditions (Prescott et al. 2003; Postgate 1968).
covers a stagnant layer that is cold and oxygen-free. GSB grow Characteristics of some of these microorganisms are presented
in a narrow horizontal band, situated just within the anaerobic in Table 3.

Table 2. Examples of energy sources for representative chemotrophs (Prescott et al. 2003).

Bacteria Electron donor Electron acceptor Carbon source Products

Thiobacilllus sp. (general) S0, H2S, S2O32- O2 CO2 SO42-

Thiobacilllus denitrifcans S0, H2S, S2O32- O2, NO3- CO2 SO42-, N2
Thiobacilllus ferrooxidans Fe2+, S0, H2S O2 CO2 Fe3+, SO42-

Volume 48 2006 CANADIAN BIOSYSTEMS ENGINEERING 2.3

2.4

Table 3. Characteristics of some microorganims implicated in degradation of H2S or other sulfur compounds.
Microorganisms
Conditions Thiobacillus Thiobacillus Thiobacillus Thiobacillus Thiobacillus Thermothrix Thiothrix Thioalkalispira Thiomicrospra
ferrooxidans thiooxidans novellus thioparus denitrificans azorensis nivea microaerophila frisia
a
pH growth range - 0.5 - 6.0 5.7 - 9.0 5-9 - 6.0 - 8.5 6 - 8.5 8 - 10.4 4.2 - 8.5
Optimum pH 1.3 - 4.5 2.0 - 3.5 7.0 7.5 6.8 - 7.4 7.0 - 7.5 - 10 6.5
a
Temperature 10 - 37 10 - 37 10 - 37 - - 63 - 86 - - 3.5 - 39
growth range (°C) (thermophile)
Optimum 30 - 35 28 - 30 30 28 28 - 32 76 - 78 15 - 30 - 32 - 35
temperature (°C)
* G+C content of 56 - 59 - 67.2 62 63 39.7 44 - 55 58.9 39.6
DNA (mol%)
a
Cells type Gram-negative Gram-negative Gram-negative Gram-negative - Gram-negative Gram-negative or Gram-negative Gram-negative
LE GÉNIE DES BIOSYSTÈMES AU CANADA

Gram-variable
a b
Group - - -2 Proteobacteria -Proteobacteria -Proteobacteria -Proteobacteria -Proteobacteria -Proteobacteria -Proteobacteria
Spore formation None None None - - None - - -
Motility 0 to several polar - Non-motile Motile Motile by means Motile No flagella Motile by means Motile
or petrichous of a polar of a single polar
flagella flagellum flagellum
Shape Rod, 0.5-1.0 m Rod, 0.5 x Rod, 0.4-0.8 x Rod, 0.9-1.8 m Rod, 0.5 x Rod, 0.3-0.8 x Rod, 0.7-2.6 x Spirillum, 0.3-0.45 Bent-rod, 0.3-0.5
1.1-2.0 m 0.8-2.0 m 1.0-3.0 m 2-5 m 0.7-5.0 m x 1-4 m x 1.0-2.7 m
Trophy Obligate Obligate Mixotroph Obligate Obligate Obligate Mixotroph Obligate Obligate
chemoautotroph chemoautotroph (faculative chemoautotroph chemoautotroph chemoautotroph (faculative chemoautotroph chemoautotroph
chemoautotroph) chemoautroph)
Examples of Ferrous ion and Hydrogen sulfide, Hydrogen sulfide, Thiosulfate, Thiosulfate, Thiosulfate, Inorganic sulfur Sulfide, Thiosulfate,
energy source reduced sulfur polithionates, methyl mercaptan, sulfide tetrathionate, tetrathionate, compounds, polysulfide, tetrathionate,
compounds elemental sulfur dimethyl sulfide, thiocyanate, hydrogen sulfide, simple organic elemental sulfur, sulfur, sulfide
dimethyl disulfide sulfide, elemental elemental sulfur compounds, sugars thiosulfate
sulfur
a
Oxygen Facultative Strictly aerobe Strictly aerobe Strictly aerobe Faculataive Strictly aerobe Strictly aerobe and Strictly aerobe and Strictly aerobe
requirement anaerobe** anaerobe*** microaerophile microaerophile
Sulfur deposit - - - Extracellular - Intracellular Intracellular Intracellular Extracellular
Reference Colorado School Takano et al. Cha et al. (1999) Vlasceanu et al. Kelly and Wood Odintsova et al. Howarth et al. Sorokin et al. Brinkhoff et al.
of Mines¶ (1997) a
Kelly et al. (2000) (1997) (2000) (1996) (1999), bPrescott et (2002) (1999)
al. (2003)
* G = guanine; C = cytosine
** Under anaerobic conditions, T. ferroosidans can grow on elemental sulfur using ferric iron as an electron acceptor.
*** Grows as an anaerobic chemoautroph by using nitrate, nitrite, or nitrous oxide.
SYED et al.

¶
http://www.mines.edu/fs_home/jhoran/ch126/thiobaci.htm (2006/02/16)
Table 4. Reactions involving chemotrophic bacteria.

Bacteria Reaction mechanism Reference

Thiobacillus 2HS- + O2 → 2S0 + 2OH- Chung et al. (1996)

thioparus 2S0 + 3O2 + 2OH- → 2SO42- + 2H+

H2S + 2O2 → SO42- + 2H+ Kim et al. (2002)

Thiobacillus 3HS- + 3.9NO3- + 0.2NH4+ + HCO3- + 1.7H+ → CH1.8O0.5N0.2 + 1.9N2 + 3SO42- + 2.3H2O Kleerebezem and
denitrificans 14.5HS- + 5NO3- + 0.2NH4+ + HCO3- + 20.3H+ → CH1.8O0.5N0.2 + 2.5N2 + 14.5S + 27.4H2O Mendez (2002)

55S + 20CO2 + 50NO3- + 38H2O + 4NH4+ → 4C5H7O2N + 25N2 + 55SO42- + 64H+ Lampe and Zhang
(1996)

5HS- + 8NO3- + 3H+ → 5SO42- + 4N2 + 4H2O McComas and

Sublette (2001)

Thiobacillus 2FeSO4 + H2SO4 + 1/2 O2 → Fe2(SO4)3 + H2O Mesa et al. (2002)

ferrooxidans
2FeS2 + 7.5O2 + H2O → Fe2(SO4)3 + H2SO4 Takano et al. (1997)

Thiobacillus sp. is widely used in studies of the conversion Other species are able to degrade sulfur compounds in
of H2S and other sulfur compounds by biological processes neutrophilic, alkaline, or thermophilic conditions. Thermothrix
(Oyarzún et al. 2003; Cha et al. 1999; Chung et al. 1996; azorensis and Thiothrix nivea are neutrophilic bacteria and
Sublette and Sylvester 1987). These bacteria have the ability to develop well at pH of 6-8 (Odintsova et al. 1996; Howarth et al.
grow under various environmental stress conditions such as 1999). Optimum growth temperature for Thermothrix azorensis,
oxygen deficiency, acid conditions, etc. Many Thiobacillus sp. a thermophilic bacterium, is between 76 and 86ºC (Odintsova et
(i.e. T. thiooxidans, T. ferrooxidans) have acidophilic character- al. 1996). Thioalkalispira microaerophila is able to grow in
istics and are able to develop in conditions of low pH (1-6). alkaline conditions and attains optimum growth at pH 10
Thiobacillus thiooxidans has a great tolerance for acidic (Sorokin et al. 2002).
conditions and can grow at pH < 1 (Devinny et al. 1999; Takano
et al. 1997). Thiobacilli such as T. thiooxidans and T. ferro- Beggiatoa sp. and Thiothrix sp., both microaerophilic -
oxidans are used in processing digested sludge or leaching low- Proteobacteria, have mixotrophic nutritional function because
grade metal ores because of their ability to remove metals by they are able to degrade H2S using inorganic and organic (e.g.
microbial leaching (Sreekrishnan and Tyagi 1996; Prescott et al. acetate) energy sources (Prescott et al. 2003; Howarth et al.
2003). Other Thiobacillus sp. (e.g. T. thioparus, T. 1999). Pseudomonas acidovorans and Pseudomonas putida are
denitrificans, T. novellus) develop in neutral medium other mixotrophs which degrade both H2S and organosulfur
(neutrophilic bacteria) at pH of 6-8. Thiobacillus denitrificans compounds (Oyarzún et al. 2003; Chung et al. 2001).
is able to grow facultatively on reduced sulfur compounds by The reaction shown in Eq. 3 takes place in an aerobic sulfide
reducing nitrate (NO3-) to nitrogen gas (N2) (Kleerebezem and removal system (Kuenen 1975).
Mendez 2002; Lampe and Zhang 1996). Thiobacillus novellus chemotrophs
is a mixotroph Thiobacilli because it can grow heterotrophically H 2 S + CO2 + nutrients + O2  →
(Kelly at al. 2000).
cells + S and / or SO42 − + H 2 O (3)

Under oxygen limiting conditions, sulfur is the major end

product, while sulfate is formed when sulfide is limited. Other
relevant reactions are shown in Table 4.
Bioreactors for H2S removal involving phototrophic
bacteria
Gas-fed batch reactor Typically a gas fed batch reactor
(Fig. 3) is a stirred tank type reactor, continuously or
intermittently operated for the gas phase (the target flux) and
cyclically operated for the liquid phase (nutritive solution). The
microorganisms can be suspended in the solution or
immobilized on different media (i.e. strontium alginate beads,
Kim and Chang 1991).
Maka and Cork (1990) investigated the effects of light
Fig. 3. Fed-batch or continuous flow reactor. quantity and intensity on H2S metabolism in a gas fed-batch

Volume 48 2006 CANADIAN BIOSYSTEMS ENGINEERING 2.5

 separation of elemental sulfur
from the bioreactor contents is
essential to realize its value as
a chemical industry feedstock.
Kim et al. (1991) tried to
optimize light energy input by
using light emitting diodes
(LEDs) with a peak wavelength
at 710 nm in a batch-fed stirred
tank reactor. Experiments were
performed separately using an
incandescent bulb, LED710, and
a combination of LED710 and a
(a) (b) fluorescent bulb as the light
sources and their individual
performances (H2S removal per
Fig. 4. Phototube reactors: (a) horizontal; (b) vertical.
unit luminous flux, (mmol/min)
per (g protein/L) per (W/m2)) were evaluated. They found that
reactor utilizing GSB. In this experiment, the intensity of the the maximum performances using LED710 and LED710 with
light source was varied, keeping the exposed area of the reactor fluorescent bulb were 18.7 and 14.1 times, respectively, the
constant at loading rates of 32 or 64 mg S2- h-1 L-1. A high performance of a reactor with an incandescent bulb.
irradiance resulted in sulfide removal, over half of which was
Later, Kim et al. (1996) investigated the performance of
converted to sulfate. At low irradiance, sulfide was not fully
LEDs in a plate type photo-bioreactor. They observed that the
oxidized and accumulated in the reactor. Between these limits,
maximum performance per unit luminous flux while using LEDs
a state in which neither sulfide nor sulfate was found in the
was 31 times that of an incandescent bulb. This efficiency was
reactor, elemental sulfur and thiosulfate (S2O32-) were the only
achieved by only supplying light within the wavelength range
products. For white light (380 to 900 nm) this optimum state
where absorption by bacteria was at a maximum.
resulted in 60% of the influent H2S becoming S0 and 40%
becoming S2O32-. For infrared light (700 to 900 nm), the Continuous-flow reactor In a continuous flow reactor (Fig. 3)
optimum favored sulfur production was 97:3 (S0: S2O32-). The the influent (gas or liquid) is continuously fed into the reactor at
optimum for infrared light also occurred at a lower irradiance the same rate it is removed from the reactor. The reactor’s
level (219 W/m2) than white light (406 W/m2). These findings contents are stirred continuously.
demonstrated the superiority of infrared light as a light source Using the effluent of a continuous-flow stirred-tank
to produce elemental sulfur using green sulfur bacteria. bioreactor, Henshaw et al. (1997) tested separation of the sulfur
Kim and Chang (1991) studied the bioconversion of by settling, settling at elevated pH, filtration, and centrifugation.
hydrogen sulfide to elemental sulfur by C. limicola in an Centrifugation produced the best separation results; 90% of the
immobilized-cell reactor and a sulfur settling tank with a free elemental sulfur and 29% of the bacteria were removed from the
cell recycle reactor. In the first reactor, cells were immobilized suspension. They noted that a continuous-flow suspended-
in strontium alginate beds and in the second reactor, the growth bioreactor system for sulfide removal/sulfur recovery
produced sulfur was removed by gravity settling and the required two separation stages, one to separate S0 from the
medium was recycled to the fed-batch reactor. In comparison bioreactor effluent and one to separate biomass from the liquid
with the free cells, the immobilized cells required 30% less light product of the first separator. A fixed-film reactor can eliminate
energy at a H2S removal rate of 68 mg h-1 L-1 initially but after or lessen the need for two separators since the biomass remains
40 hours, the deterioration of the H2S removal efficiency in the reactor.
became significant due to the accumulation of sulfur in the beds.
Subsequently, Kim et al. (1992) compared sulfide removal rates Phototube reactors Two types of phototube reactors are
in 2-L and 4-L reactors. The difference in sulfide removal rates shown in Figs. 4a and 4b. These are tubular type reactors that
between 2-L and 4-L reactors (0.11 and 0.07 mg H2S/h per are continuously operated. The reactor can be horizontally
milligram of protein, respectively) was explained by the higher oriented (Fig. 4a) having several passes or spirals to improve the
light attenuation in the larger reactor since light intensity residence time in the reactor (Kobayashi et al. 1983) or can be
decreased exponentially with the penetration depth. They also vertically oriented, as presented in Fig. 4b (Syed and Henshaw
observed that the average diameter of sulfur aggregates was 10 2003; Henshaw et al. 1999). The material of the tube is
times that of bacterial cells. transparent to light and impermeable to oxygen (Syed and
Henshaw 2003). Bacteria develop on the inner wall of the tube
Basu et al. (1996) used a continuous stirred tank reactor reactor (fixed-film reactor).
equipped with a sulfur separator to remove hydrogen sulfide
from a gas stream containing 2.5% H2S at 1 atmosphere Kobayashi et al. (1983) used a “phototube” reactor in which
pressure. At a sulfide loading rate of 94.4 mg h-1 L-1, H2S a sulfide containing reactor was passed through a 12.8 m long,
conversion by Chlorobium thiosulfatophilum ranged from 53% 3.2 mm ID Tygon tube which was immersed in an illuminated
at a gas retention time of 12.2 min to 100% at a gas retention water bath. The tube was able to achieve 95% sulfide removal
time of 23.7 min. The sulfur recovered from the process by in about 24.6 min while operated at a sulfide loading rate of
gravity separation was 99.2% of the theoretical yield. The 67 mg h-1 L-1.

2.6 LE GÉNIE DES BIOSYSTÈMES AU CANADA SYED et al.

 Postulating that a vertical attached growth configuration loading rate, the stirred-tank reactor’s removal efficiency was
could eliminate or significantly reduce the problem of sulfur approximately 62%.
accumulation encountered by Kim and Chang (1991), Henshaw
Bioscrubbers Removal of H2S using bioscrubbers (Fig. 5a)
et al. (1999) tested the ability of GSB to grow on six different
involves a two stage process, firstly absorption of H2S by a
transparent plastic tube materials to determine which medium
liquid followed by biological oxidation of H2S in the liquid.
would be best as a support medium in a vertical flow fixed-film
bioreactor. Tygon was found superior to the other materials Nishimura and Yoda (1997) used a multiple bubble-tray air-
tested in terms of number of bacteriochlorophyll per unit surface tight contact tower (bioscrubber) to scrub hydrogen sulfide from
area and low oxygen permeability. Bacteriochlorophyll was the biogas produced by an anaerobic wastewater treatment
used as a measure of GSB biomass. process. A two-reactor system (a gas-liquid contact tower and
an aeration tank) were used to separate the oxidation process
Using a fixed film, continuous-flow photobioreactor
from the absorption process to prevent air from mixing with the
employing C. limicola and a infrared light source, Henshaw and
biogas. Mixed liquor from the activated sludge process was
Zhu (2001) succeeded in removing sulfide from synthetic
continuously fed to and withdrawn from the contact tower. In
wastewater at a sulfide loading rates of 111 to 286 mg h-1 L-1,
the contact tower, H2S from the biogas was absorbed into the
while 92 to 95% of the influent sulfide was converted to
mixed liquor and subsequently oxidized to sulfate by sulfur
elemental sulfur. A fixed-film reactor was selected because of
oxidizing bacteria after returning to the aeration tank. Based on
its ability to retain biomass for further sulfide oxidation. In this
their preliminary results, a full scale plant treating potato
process, light can easily be transmitted to the biomass adhering
processing wastewater was constructed. When treating
to the inside of the transparent tubes whereas in a suspended
2000 ppm of H2S in 40 m3/h of biogas, more than 99% removal
growth process, light can be considerably attenuated at the
efficiency was achieved.
center of the reactor. In subsequent experiments, Syed and
Henshaw (2003, 2005) investigated the effects of tube size and Mesa et al. (2002) described a bioscrubber system which can
light quality on H2S removal. They observed that the smallest be integrated into a system to remove H2S from biogas by a
diameter tube reactor achieved the highest sulfide removal at the combination of chemical and biological processes. H2S removal
same light intensity (of infrared bulb and LEDs). A higher can be achieved by absorption in a ferric sulfate solution
sulfide loading rate was achieved when LEDs providing light producing ferrous sulfate and elemental sulfur. Ferric sulfate can
matching the peak absorption spectrum of GSB were the light be regenerated by biological oxidation using Acidithiobacillus
source. The reactor system can be used for removal of gaseous ferrooxidans. Relevant reactions are shown in Table 4. The
H2S after dissolving it in water. study investigated the oxidation of ferrous iron by A.
ferrooxidans which was immobilized on a polyurethane foam
Bioreactors for H2S removal involving chemotrophic support and the support particles placed in an aerated column.
bacteria Ferric precipitates were accumulated on the support and on the
A number of studies, summarized in Table 5, have been air diffusers which necessitated periodic interruptions of the
conducted using chemotrophic bacteria. process for cleaning. Precipitation, air supply, and chemical cost
are the potential constraints for this process.
Gas-fed batch reactors Janssen et al. (1995) used two batch-
fed reactors to study the oxidation of sulfide using a mixed A full scale plant located northeast of Brooks, Alberta,
culture of Thiobacilli. Pure oxygen was supplied to the reactors. Canada uses Shell-Paques® process for natural gas
The maximum sulfur production (73 ± 10%) occurred at an desulfurization (Benschop et al. 2002). H2S is removed from a
oxygen to sulfide ratio of 0.6 to 1.0 mol L-1 h-1/ mol L-1 h-1. At gaseous stream by absorption into a sodium
lower oxygen to sulfide ratios, the lower biological oxidation carbonate/bicarbonate solution. The sulfide containing
capacity resulted in the production of more thiosulfate. At scrubbing liquid is treated in the bioreactor where it is mostly
higher oxygen to sulfide ratios, more sulfate was produced converted biologically to elemental sulfur. The bioreactor is
because more energy was consumed for bacterial growth than supplied with a nutrient stream, air, make-up water, and sodium
for the formation of elemental sulfur. hydroxide. It is reported that normally less than 3.5% of the
sulfide is converted to sulfate and a continuous bleed stream is
Continuous-flow reactors Sublette and Sylvester (1987) required to avoid accumulation of sulfate. A compost filter is
reported on a continuous stirred-tank reactor (CSTR) system used to treat the trace H2S present in the spent air from the
using Thiobacillus denitrificans to remove H2S from gas bioreactor. Less than 4 ppmv effluent H2S concentration is
streams. Ninety-seven percent of the H2S bubbled was removed achieved when treating natural gas containing 2000 ppmv H2S.
and oxidized to sulfate.
Biofilter A biofilter (Fig. 5b) is a three phase bioreactor (gas,
Buisman et al. (1990) tested three different continuous-flow
liquid, solid) made with a filter bed that has a high porosity,
reactor configurations: fixed-film CSTR, biorotor (a rotating
high buffer capacity, high nutrient availability, and high
cage containing reticulated polyurethane biomass support
moisture retention capacity to ensure that the target
particles, partly immersed in the reactor liquid), and a fixed-film
microorganisms can grow on it (Elias et al. 2002; Jorio and
upflow reactor. For the upflow and biorotor reactors, 95 to
Heitz 1999; Daustos et al. 2005). The contaminated gas is
100% sulfide removal efficiencies were achieved for loading
continuously fed in the biofilter, while a nutrient solution is
rates up to 500 mg H2S h-1 L-1. The removal efficiency decreased
discontinuously added. Various types of biofilter media have
rapidly above this loading rate. At 938 mg h-1 L-1 (biorotor) and
been used by researchers. Representative cases are discussed
1040 mg h-1 L-1 (upflow) loadings, sulfide removal efficiencies
below.
were 69 and 73%, respectively. At a 500 mg h-1 L-1 sulfide

Volume 48 2006 CANADIAN BIOSYSTEMS ENGINEERING 2.7

2.8
Table 5. Research conducted on hydrogen sulfide removal using bioscrubbers/biofilters or biotrickling filters.

Gas flow rate

Pollutants treated
and empty bed Removal
Scale Process type Type bed Bed volume and inlet gas Bacteria Innoculation Reference
residence time efficiency
concentration
(EBRT)
Laboratory Biofiltration Cell-laden 0.7 L 5-100 ppm H2S, 18-150 L/h T. thioparus 85-99% (H2S) Immobilization of Chung et al.
Ca-alginate balance air T. thioparus on (1996)
Ca-alginate
Full Bioscrubbing: Activated sludge 3 m3 Biogas from waste- 40 m3/h Indigenous >99% (H2S) - Nishimura and
anerobic absorption + (scrubber) water treatment plant: (Thiobacillus sp., Yoda (1997)
aerobic biooxidation 550 m3 300-2000 ppm H2S, etc.)
(aeration tank) 80% CH4, 20% CO2

Laboratory Biooxidataion in the Poly-4- 6.3 L (reactor Mixed gas composed - T. novellus 100% (H2S) Immobilization of Cha et al.
bubble column reactor vinylpyridine volume) of 30 ppm H2S, MM, 100% (MM) T. novellas on PVP (1999)
(PVP) matrix DMS, and DMDS* 87% (DMDS) matrix)
LE GÉNIE DES BIOSYSTÈMES AU CANADA

73% (DMS)
Laboratory Biofiltration Compost or 18 L 10-450 ppm H2S, 1.7 m3/h Indigenous 90-100% (H2S) Yes (waste- Wani et al
compost/hog fuel balance air 38s (in sludge) activated sludge) (1999)

10-450 ppm H2S, 90-100% (H2S)

10.8 ppm DMS, 30-35% (DMS)
6.6 ppm DMDS, <30% (DMDS)
balance air
Laboratory Biofiltration Cell-laden Ca- 0.7 L 60-120 ppm H2S, 36 L/h Pseudomonas >90% (H2S) Co-immobilization Chung et al.
alginate 60-120 ppm NH3, 72 s putida (for H2S) >95% (NH3) of P. putida and (2001)
balance air Arthrobacter A. oxydans on
oxydans (for NH3) Ca-aglinate
Laboratory Biotrickling Polypropylene 10 L 170 ppm H2S, 1 m3/h Toluene-degraders 100% (H2S) Yes (biomass from Cox and
pall rings 2.2 g/m3 toluene, 36 s sulfide oxidizing 25-75% a toluene- Deshusses
balance air bacteria (toluene) degrading (2001)
biotrickling filter)
Laboratory Biofiltration Pig manure + 5.9 L 10-45 g H2S m-3 h-1, 13.5-27 s Indigenous >90% (H2S) No Elias et al.
sawdust balance air (2002)
Pilot Biotrickling - 3.8 m3 Emissions from 650 m3/h - 98% (H2S) - Cox and
wastewater treatment 21 s 50-70% (VOC) Deshusses
plants: 10-50 ppm (2002)
H2S, 0-150 ppb VOC,
traces of other
compounds
Full Biotrickling Structured plastic 51 m3 Exhaust air from a 44,200 m3/h - 85-99% (H2S) - Cox and
packing cellophane plant: (total flow) 40-70% (CS2) Deshusses
60-155 ppm H2S, 4-10 s (2002)
SYED et al.

35-100 mg/m3 CS2

Volume 48
Table 5. continued

Full Biotrickling/ Polyurethane 500 m3 Odors from tobacco 11 s - >90% (H2S) - Cox and
Biofiltration foam (6 units) company: 800-1200 OU Deshusses
(2002)
2006

Laboratory Biofiltration Wood chips, 1L 30-450 ppm H2S, 60-180 T. thioparus 75-99% (H2S) Yes (activated Kim et al.
granular 35-200 ppm NH3, L/h Nitrifying 30-92% (NH3) sludge from (2002)
activated carbon balance air 20-60 s bacteria sewage water
(GAC) treatment plant
+ culture of
T. thioparus)
Laboratory Biooxidation in the Polyurethane 1L Effluent from a gas lift - T. >80% (S-tot) - Kleerebezem
fixed film reactor particles reactor treating a sulfide tenitrificans and Mendez
nitrate mixture Nitrifying (2002)
(2-20 mM as S-tot) bacteria
Full Biosscrubbing Alkaline solution - Natural gas: 322,000 Indigenous >99.8% (H2S) - Benschop et
(absorption +aerobic for scrubbing 2000 ppm H2S nm3/d al. (2002)
CANADIAN BIOSYSTEMS ENGINEERING

biooxidation) + Compost for

separate biofiltration biofiltration
Full Biofiltration Wood-based Open bed Odorous air from meat 25485 Indigenous >96% (all No Shareefdeen
medium biofilter rendering plant (24,544 m3/h compounds, et al. (2002)
OU): 1.07 mg H2S/m3, 5.2 30 s with exception
mg NH3/m3, 0.66 mg/m3 of methanethiol,
methanethol, 1.2 mg/m3 70%)
ethylamine, 775.25 mg/m3
DMS
Full Biotrickling Polyurethane 7.3 m3 Odorous air from 1.6-2.3 s Thiobacillus >97% (H2S) Yes (activated Gabriel and
foam wastewater treatment sp. 67% (MM) sludge from Deshusses
plant: 5-25 ppm H2S, 67 44% (carbonyl wastewater (2003)
ppb carbonyl sulfide, 192 sulfide) treatment
ppb MM, 70 ppb CS2 35% (CS2) plant)
(4000 ppm CO2)
Laboratory Biofiltration Granulated 10 L 170-680 g H2S m-3 d-1, 86- - Nitrifying 100% (H2) Yes Malhautier et
sludge (column) 340 g NH3 m-3 d -1, bacteria, 80% (NH3) (acclimatized al. (2003)
balance air Sulfide sludge)
oxidizing
bacteria
Laboratory Biofiltration Mature compost 8L 50 ppm H2S, 10 L/min Indigenous 90-100% (H2S) No Morgan-
balance air Sagastume et
al. (2003)
Laboratory Biofiltration Peat 1L 355-1400 ppm H2S, 30 L/h T. thioparus 65-100% (H2S) Yes (culture of Oyarzún et
balance air T. thioparus) al. (2003)
Full Biofiltration - - Biogas from anaerobic 10-350 Thiobacteria >90% (H2S) Yes (activated Schieder et
wastewater treatment m3/h (Thiobacillus sludge from al. (2003)
plant (5000 ppm H2S) sp.) digester)
* MM = methyl mercaptan; DMS = dimethyl sulfide; DMDS = dimethyl disulfide
2.9
 (a) (b) (c)
Fig. 5. Systems for removal of H2S: (a) bioscrubber; (b) biofilter; (c) biotrickling filter.

Chung et al. (1996) immobilized Thiobacillus thioparus A comparison between removal efficiencies of inorganic
CH11 with Ca-alginate producing pellet packing material for the (H2S) and organo-sulfur (methyl mercaptan, dimethyl sulfide,
biofilter. At a 28 second optimal retention time, the H2S and dimethyl disulfide) odour compounds by immobilized T.
removal efficiency was more than 98%. Elemental sulfur or novellus is presented in the study of Cha et al. (1999). They
sulfate was produced depending on the inlet H2S concentration. observed T. novellus can degrade H2S > methyl mercaptan >
dimethyl disulfide > dimethyl sulfide and the removal efficiency
Chung et al. (1997) used Thiobacillus novellus in a biofilter for
was 100% for H2S and methyl mercaptan, 87% for dimethyl
H2S oxidation under mixotrophic conditions. A removal disulfide, and 73% for the dimethyl sulfide. The final metabolic
efficiency of 99.6% was achieved and the products were sulfate product was sulfate.
(83.6%) and sulfite (12.6%). Little conversion of sulfide to
Shareefdeen et al. (2002) reported the operation of a
elemental sulfur was achieved. Later, Chung et al. (2001) used commercial biofilter for the treatment of an air stream
biofilters packed with co-immobilized cells Pseudomonas containing hydrogen sulfide, ammonia, dimethyl sulfide,
putida CH11 and Arthobacter oxydans CH8 for removal of H2S methanethiol, and ethylamine. This proprietary wood-based
and NH3, respectively, which are often present in off-gases of a (BIOMIX™) biofilter achieved 96.6% removal of H2S at an
livestock farm. In the 5-65 ppm range, H2S and NH3 removal inlet concentration of 1.07 mg/m3.
efficiencies were greater than 96%. However, at higher Elias et al. (2002) used packing material made up of pig
concentrations, H2S and NH3 showed inhibitory effects on H2S manure and sawdust for biofiltration purposes. More than 90%
removal. They also assessed the environmental risk associated H2S removal efficiency was attained at a loading rate of
with the release of bacteria when treating large volumes of 45 g m-3 h-1. No nutrient was added to the system and the
waste gases. The exhaust gas contained small amounts of porosity of the packing material decreased from 23.1 to 12.9%.
bacteria (< 19 CFU/m3 in all cases) and was considered safe. However, this change in porosity did not affect the removal
efficiency significantly and it was claimed that the biofilter
Wani et al. (1999) described the removal characteristics of could be easily cleaned by flushing water through the inlet. The
H2S and other reduced sulfur compounds emitted from kraft main by-product of the biodegradation process was sulfur (82%
pulp mills using three different biofilter mediums: compost, of total sulfur accumulation), accompanied by sulfates and
hog-fuel (pulverized mixture of raw bark, wood waste, and other thiosulfates (<18%).
materials) and a mixture of compost and hog fuel at 1:1 (w/w) Kim et al. (2002) investigated the simultaneous removal of
ratio. Dolomitic lime was mixed with each medium to act as a H2S and NH3 using two biofilters, one packed with wood chips
pH buffer. No significant difference was observed in the H2S and the other with granular activated carbon (GAC). A mixture
elimination capacities of these three media. However, the pH of of activated sludge (as a source of nitrifying bacteria) and
the media decreased significantly over an operating period of Thiobacillus thioparus (for sulfur oxidation) was sprayed on the
more than six months. At H2S concentrations up to 250 ppmv, packing materials and the drain solution of the biofilter was
complete removal was observed. The removal efficiency for recirculated to increase the inoculation of microorganisms.
inlet concentrations higher than 250 ppmv was above 90%. Initially both of the filters showed high (99.9%) removal
Compost, hog-fuel, and the mixture media had maximum elim- efficiency. However, due to the accumulation of elemental
ination capacities of 136, 137, and 138 g m-3 h-1, respectively. sulfur and ammonium sulfate on the packing materials removal

2.10 LE GÉNIE DES BIOSYSTÈMES AU CANADA SYED et al.

efficiency decreased over time to 75 and 30% for H2S and NH3, Biotrickling filter. The working principle of a biotrickling
respectively. filter (Fig. 5c) is the same as for a biofilter except that the
Kleerebezem and Mendez (2002) proposed the simultaneous packed bed is continuously trickled over by an aqueous phase
degradation of H2S and NH3 using T. denitrificans and nitrifying nutritive solution (Cox and Deshusses 2001).
bacteria. Thiobacillus denitrificans is able to degrade H2S in Cox and Deshusses (2001) used two laboratory scale
both aerobic and anaerobic conditions using oxygen or nitrate biotrickling filters made of polypropylene, inoculated with
as an electron acceptor. Nitrate (NO3-) can be obtained from the biomass from a toluene biodegrading filter operating at pHs of
nitrification of ammonia by nitrifying bacteria. Thiobacillus 7.0 and 4.5 to treat H2S and toluene in a gas stream. There was
denitrificans simultaneously oxidizes H2S to elemental sulfur or no significant difference between the performances of the two
sulfate and reduces nitrate to nitrogen gas. Stoichiometric reactors in terms of H2S removal. At an inlet concentration of
equations for autotrophic denitrification by T. denitrificants are approximately 50 ppmv, complete consumption of H2S was
presented in Table 4 (Kleerebezem and Mendez 2002; Lampe observed. However, the removal efficiency decreased to 70-
and Zhang 1996). 80% when the inlet concentrations were raised to 170 ppmv.
Malhautier et al. (2003) also performed an experiment High removal efficiency for H2S, in comparison to other
involving H2S and NH3 using two laboratory scale biofilters reduced sulfur compounds was obtained by Gabriel and
packed with granulated digested sludge. One unit was fed Deshusses (2003) (Table 5) using Thiobacillus sp. in a
mainly with H2S and the other unit with NH3. Complete H2S biotrickling filter. For inlet H2S concentrations as high as
removal (100%) was obtained and no influence on NH3 or H2S 30 ppmv, typical removal efficiency was 98%. Methyl
removal was observed. An 80% NH3 removal efficiency was mercaptan, carbonyl sulfide, and carbon disulfide removal
obtained, however, the authors concluded that the oxidation of efficiencies were 67, 44, and 35% at inlet concentrations of 67,
high levels of H2S might have a negative effect on the growth 193, and 70 ppbv, respectively.
and activity of nitrifying bacteria. Sercu et al. (2005) studied the aerobic removal of hydrogen
Morgan-Sagastume et al. (2003) investigated changes in the sulfide using a biotrickling filter packed with 1L-polyethylene
physical properties of a compost biofilter treating hydrogen rings (73% volume free) inoculated with Acidithiobacillus
sulfide. Bench-scale biofilter columns filled with compost media thiooxidans ATCC-19377. The inlet H2S concentration was
consisting of mature compost derived from food, leaf, yard varied between 400 and 2000 ppm and the airflow rate was
varied between 0.03 and 0.12 m3/h. However, the system
waste, and horse manure were used. H2S removal efficiency
performance was not affected by changing the operational
decreased from 100 to 90% over 206 days of operation. They
conditions and a maximal removal efficiency of 100% was
concluded that the variation in moisture content and specific
obtained. During the experiment, the pH of the nutritive
surface area can explain the decrease in removal efficiency over solution decreased to 2-3, but this did not affect the process
time. They also mentioned that SO42- accumulation which performance.
reduced the pH of the compost media from 7.5 to 4.5 could be
an important factor. After re-mixing of the compost media, the Soreanu et al. (2005) developed a laboratory-scale
H2S removal efficiency returned to near 100%. biotrickling system in order to remove H2S from digester biogas
under anaerobic conditions. In these experiments,
Oyarzún et al. (2003) used peat for the filter bed of a polypropylene balls inoculated with anaerobically digested
biofiltration system inoculated with Thiobacillus thioparus. sludge were used as packing material in the bioreactor (packing
Supplemental nutrient was added and the initial moisture volume of 0.0062 m3, 90% volume free). Sodium sulfite was
content was adjusted to 92%. The pH was also adjusted to 6.0. added in the nutritive solution as an oxygen scavenging agent.
Full removal was achieved when fed with 355 ppm H2S at Nitrate was used as electron acceptor in the absence of oxygen.
0.03 m3/h. The removal efficiency decreased with increasing Removal efficiency greater than 85 % was achieved for an H2S
inlet H2S concentrations and a maximum removal capacity of inlet concentration of 500 ppm and a gas flowrate of 0.05 m3/h.
55 g m-3 h-1 was obtained. Of particular interest, inhibition of the biological process by
Schieder et al. (2003) described the “BIO-Sulfex” biofilter trace amounts of O2 was noticed when a nitrate solution was
to remove H2S from biogas which uses thiobacteria attached on used as the sole nitrogen/nutrient source.
fixed bed material. The biomass was aerated and the filter was
flushed with nutrient containing liquid to remove sulfur from the CONCLUSION
system. Six BIO-Sulfex modules to treat biogas containing up The above discussion shows that the preferred treatment method
to 5000 ppm H2S were operated at flowrates of 10 to 350 m3/h for H2S gas containing streams depends on the source of the
with 90% or more H2S removal achieved. gas. In the case of H2S in biogas, anaerobic methods involving
Clark et al. (2004) operated a pilot-scale agricultural phototrophic bacteria provide the inherent advantage of
biofilter to reduce odours from a swine manure treatment plant’s maintaining the anaerobic nature of the gas and avoid any
exhaust air. Biofilters were packed with polystyrene particles potential safety problems. Cholorobium limicola is a desirable
and peat moss (3:1 ratio by volume). The packing volume was bacterium to use due to its growth using only inorganic
1.89 m3. The gas flowrate was 100 L/s. The inlet load contained substrates, efficiency at converting sulfide to elemental sulfur,
2-60 ppm hydrogen sulfide and 2-30 ppm ammonia. The and extracellular production of elemental sulfur. Because of the
addition of nutrients did not play an important role in the overall reduced requirement for separation of the biomass and the
system performance. Average odour reduction was ability to achieve higher loadings, fixed-film reactors have the
approximately 38% without addition of nutrients and 45% when greatest potential for more cost-effective sulfide conversion. If
nutrients were added. Increasing the temperature had a light is controlled properly, the oxidation product is mostly
favourable effect during the acclimatisation phase only. sulfur which is non toxic and non-corrosive. Light supply is one

Volume 48 2006 CANADIAN BIOSYSTEMS ENGINEERING 2.11

of the key cost components and the major constraint for this ACKNOWLEDGMENTS
process. It is important to provide economical light energy so
that the process can be established as a viable alternative to The authors thank Environment Canada and Natural Resources
physicochemical processes currently in use. The development Canada for financing this work and the National Science and
and use of energy efficient LEDs and the proper use of sunlight Engineering Research Council of Canada for administration of
and reflectors to minimize the electrical energy use are the the first two authors’ visiting research fellowships.
current research needs.
REFERENCES
Chemotrophic bacteria require careful control of the oxygen
concentration to produce elemental sulfur instead of sulfate. Anderson, J.W. 1978. Sulfur in Biology. London, UK: Edward
Even at the highest rates of conversion of sulfide to elemental Arnold (Publishers) Limited.
sulfur, some sulfate is still produced. Bacteria of Thiobacillus Basu, R., E.C. Clausen and J.L. Gaddy. 1996. Biological
sp. which have the ability to grow under various environmental conversion of hydrogen sulfide into elemental sulfur.
stress conditions such as oxygen deficiency, acid conditions, etc. Environmental Progress 15(4): 234-238.
are widely used for the conversion of H2S and other sulfur
compounds by biological processes. The removal of H2S using Benschop, A., A. Janssen, A. Hoksberg, M. Seriwala, R. Abry
bioscrubbers involves a two stage process: absorption of H2S by and C. Ngai. 2002. The shell-Paques/THIOPAQ gas
a liquid medium followed by the biological oxidation of H2S in desulphurization process: Successful start up first
the liquid. Little data on the cost of plants using bioscrubbers is commercial unit. http://www.paques.nl (2006/02/15)
available, however the existence of full scale facilities of this Boyd, R. 2000. Internalising environmental benefits of
nature are a good indication of their cost-effectiveness. anaerobic digestion of pig slurry in Norfolk. Student
Biofiltration has also been demonstrated for the removal of research paper. University of East Anglia, Norwich,
H2S from wastewater treatment plant off-gases. It is also Norfolk, UK..
effective in removing other volatile organic compounds present Brinkhoff, T., G. Muyzer, O.C. Wirsen and J. Kueven. 1999.
in biogases. The process is essentially aerobic and the oxidation Thiomicrospira kuenenii sp. nov. and Thiomicrospira frisia
product is mostly sulfate rather than sulfur. Drawbacks of this sp. nov., two mesophilic obligately chemolithoautotrophic
technology are that the sulfate can reduce the pH of the media sulfur-oxidizing bacteria isolated from an intertidal mud flat.
and sulfur deposition can reduce its efficiency. However, the International Journal of Systematic Bacteriology 49: 385-
biofilter medium is inexpensive and may contain sufficient 392.
nutrients for the microbial communities. The process is also Buisman, C.J.N., R. Post, P. Ijspreet, S. Geraats and G.
suitable for treating off-gases from livestock farming which Lettinga. 1989. Biotechnological process for sulphide
contain a mixture of H2S and NH3. Biofiltration by aerobic removal with sulphur reclamation. Acta Biotechnologica 9:
bacteria is not suitable for the treatment of biogas since the 271-283.
presence of significant levels of oxygen in the purified gas may
pose a potential safety problem. Buisman, C.J.N., B. Wit and G. Lettinga. 1990.
Biotechnological sulphide removal in three polyurethane
Anaerobic biofiltration of H2S can provide the same carrier reactors: stirred reactor, biorotor reactor and upflow
advantages of aerobic biofiltration including the use of low cost reactor. Water Research 24(2): 245-251.
materials, the exclusion of aeration costs as well as the
elimination of safety risks associated with operation in an Cha, J.M., W.S. Cha and J.H. Lee. 1999. Removal of organo-
oxygen-rich environment. The use of T. denitrificans has great sulphur odour compounds by Thiobacillus novellas SRM,
potential in this respect and nitrified mixed liquor suspended sulphur-oxidizing microorganisms. Process Biochemistry
solids from activated sludge/biological nutrient removal plant as 34(6-7): 659-665.
suggested by Kleerebezem and Mendez (2002) can be a useful Chung, Y.C., C. Huang and C.P. Tseng. 1996. Operation
source of nitrate. Several research studies are expected in this optimization of Thiobacillus thioparus CH11 biofilter for
area. A system combining the advantages of phototrohy and hydrogen sulfide removal. Journal of Biotechnology 52: 31-
chemotrophy may also be considered. 38.
Two separate biofilters employing C. limicola and T. Chung, Y.C., C. Huang and C.F. Li. 1997. Removal
denitrificans can be used alternatively for treating biogas from characteristics of H2S by Thiobacillus novellous CH 3
a single source. Bioreactor/filter containing phototrophic biofilter in autotrophic and mixotrophic environments.
bacteria can treat hydrogen sulfide during daytime using Journal of Environmental Science and Health A32 (5):
sunlight, whereas biogas can be directed to the biofilter 1435-1450.
containing T. denitrificans during the night. Operational Chung, Y.C., C. Huang and C.P. Tseng. 2001. Biological
flexibility regarding sulfur recovery and filter cleaning can be elimination of H2S and NH3 from wastegases by biofilter
achieved using this dual filter process. packed with immobilized heterotrophic bacteria.
Finally, for a biogas purification system, simultaneous Chemosphere 43: 1043-1050.
removal of hydrogen sulfide and siloxane is required since Clark, O.G., I. Edeogu, J. Feddes, R.N. Coleman and A.
siloxane itself can cause significant damage to combustion Abolghasemi. 2004. Effects of operating temperature and
engines. Grumping et al. (1999) has reported anaerobic supplemental nutrients in a pilot-scale agricultural biofilter,
microbial degradation of siloxane. The development of an Canadian Biosystems Engineering 46: 6.7-6.16.
anaerobic process which is able to remove both hydrogen Cooper, D.C. and E.C. Alley. 1986. Air Pollution Control: A
sulfide and siloxane simultaneously is an important current Design Approach (Chapter 14: 411-450). Boston, MA:
research need. PWS Engineering.

2.12 LE GÉNIE DES BIOSYSTÈMES AU CANADA SYED et al.

Cork, D., J. Mather, A. Maka and A. Srnak. 1985. Control of Kelly, P.D. and P.A. Wood. 2000. Confirmation of Thiobacillus
oxidative sulfur metabolism of Chlorobium limicola forma denitrificans as a species of the genus Thiobacillus, in the -
thiosulfatophilum. Applied Environmental Microbioogy 49: subclass of the Proteobacteria, with strain NCIMB 9548 as
269-272. the type strain. International Journal of Systematic and
Cox, H.H.J. and A.M. Deshusses. 2001. Co-treatment of H2S Evolutionary Microbiology 50: 547-550.
and toluene in a biotrickling filter. Chemical Engineering Kelly, D.P., I.R. McDonald and A.P. Wood. 2000. Proposal for
Journal 3901: 1-10. the reclassification of Thiobacillus novellus as Starkeya
Cox, H.H.J. and A.M. Deshusses. 2002. Biotrickling filters for novella gen. nov., comb. nov., in the -subclass of the
air pollution control. In The Encyclopedia of Environmental Proteobacteria. International Journal of Systematic and
Microbiology, ed. G. Britton, 2: 782-795. New York, NY: Evolutionary Microbiology 50: 1797-1802.
J. Wiley & Sons. Kim, B.W and H.N. Chang. 1991. Removal of hydrogen sulfide
Dastous, P.A., G. Soreanu, J. Nikiema and M. Heitz. 2005. by Chlorobium thiosulfatophilum in immobilized cell and
Biofiltration of three alcohols on a mature bed compost. In sulfur settling free-cell recycle reactors. Biotechnology
2005 A&WMA Annual Conference Proceedings CD-ROM, Progress 7: 495-500.
Paper #1038. Pittsburgh, PA: Air & Water Management Kim, B.W., E.H. Kim and H.N. Chang. 1991. Application of
Association.. light emitting diodes as a light source to a photosynthetic
Devinny, S.J., A.M. Deshusses and S.T. Webster. 1999. culture of Chlorobium thiosulfatophilum. Biotechnology
Biofiltration for Air Pollution Control. Boca Raton, FL: Techniques 5(5): 343-348.
CRC/Lewis Publishers. Kim, B.W., H. N. Chang, I.K. Kim and K.S. Lee. 1992. Growth
Elias, A., A. Barona, A. Arreguy, J. Rios, I. Aranguiz and J. kinetics of the photosynthetic) bacterium Chlorobium
Peñas. 2002. Evaluation of a packing material for the thiosulfatophilum in a fed-batch reactor. Biotechnology
biodegradation of H2S and product analysis. Process Bioengineering 40: 583-592.
Biochemistry 37: 813-820. Kim, Y.J., B.W. Kim and H.N. Chang. 1996. Desulfurization in
Gabriel, D. and A.M. Deshusses. 2003. Retrofitting existing a plate-type gas-lift photobioreactor using light emitting
chemical scrubbers to biotrickling filters for H2S emission diodes. Korean Journal of Chemical Engineering 13(6):
control. In Proceedings of the National Academy of Science 606-611.
of the United States of America 100(11): 6308–6312. Kim, H., J.Y. Kim, S.J. Chung and Q. Xie. 2002. Long-term
Grumping, R., K. Michalke, A.V. Hirner and R. Hensel. 1999. operation of a biofilter for simultaneous removal of H2S and
Microbial degradation of Octamethylcyclotetrasiloxane. NH3. Air & Waste Management Association 52: 1389-1398.
Applied and Environmental Microbiology 65(5): 2276- Kleerebezem, R. and R. Mendez. 2002. Autotrophic
2278. denitrification for combined hydrogen sulfide removal from
Henshaw, P.F., J.K. Bewtra and N. Biswas. 1997. Biological biogas and post-denitrification. Water Science and
conversion of sulfide to elemental sulfur. Indian Journal of Technology 45(10): 349-356.
Engineering & Material Sciences 5: 202-210. Kuenen, J.G. 1975. Colourless sulphur bacteria and their role in
Henshaw, P.F., D. Medlar and J. McEwen. 1999. Selection of the sulphur cycle. Plant Soil 43: 49-76.
a support medium for a fixed-film green sulfur bacteria Kobayashi, H.A., M. Stenstrom and R.A. Mah. 1983. Use of
reactor. Water Research 33(14): 3107-3110. photosynthetic bacteria for hydrogen sulfide removal from
Henshaw, P.F. and W. Zhu. 2001. Biological conversion of anaerobic waste treatment effluent. Water Research 17(5):
hydrogen sulfide to elemental sulfur in a fixed-film 579-587.
continuous flow photo-reactor. Water Research 35 (15): Lampe, D.G. and T.C. Zhang. 1996. Evaluation of sulfur-based
3605-3610. autotrophic denitrification. In Proceedings of the HSRC/
Howarth, R., F.R. Unz, M.E. Seviour, J.R. Seviour, L. Blackall, WERC Joint Conference on the Environment. Great Plains/
W.R. Pickup, G.J. Jones, J. Yaguchi and I. Head. 1999. Rocky Mountain Hazardous Substance Research Center.
Phylogenetic relationships of filamentous sulfur bacteria http://www.engg.ksu.edu/HSRC/96Proceed/lampe.pdf
(Thiothrix spp. and Eikelboom type 021N bacteria) isolated (2006/02/16)
from wastewater treatment plants and description of Larsen, H. 1952. On the culture and general physiology of the
Thiothrix eikelboomii sp. nov., Thiothrix unzii sp. nov., green sulfur bacteria. Journal of Bacteriology 64: 187-196.
Thiothrix fructosivorans sp. nov. and Thiothrix defluvii sp.
nov. International Journal of Systematic Bacteriology 49: Lastella, G., C. Testa, G. Cornacchia, M. Notornicola, F.
1817-1827. Voltasio and V.K. Sharma. 2002. Anaerobic digestion of
semi-solid organic waste: Biogas production and its
Janssen, A.J.H., R. Sleyster, C.V.D. Kaa, A. Jochemsen, J. purification. Energy Conversion and Management 43: 63-
Bontsema and G. Lettinga. 1995. Biological sulphide
75.
oxidation in a fed-batch reactor. Biotechnology
Bioengineering 47: 327-333. Maka, A. and D. Cork. 1990. Quantum efficiency requirements
for an anaerobic photobioreator. Journal of Industrial
Jorio, H. and M. Heitz. 1999. Traitement de l’air par
Microbiology 5: 337-354.
biofiltration. Canadian Journal of Civil Engineering 26:
402-424.

Volume 48 2006 CANADIAN BIOSYSTEMS ENGINEERING 2.13

Malhautier, L., C. Gracian, C.J. Roux, L.J. Fanlo and P. Le Shareefdeen, Z., B. Herner and S. Wilson. 2002. Biofiltration
Cloirec. 2003. Biological treatment process of air loaded of nuisance sulfur gaseous odors from a meat rendering
with an ammonia and hydrogen sulfide mixture. plant. Journal of Chemical Technology and Biotechnology
Chemosphere 50: 145-153. 77: 1296-1299.
McComas, C. and L.K. Sublette. 2001. Characterization of a Soreanu, G., M. Al-Jamal and M. Béland. 2005. Biogas
novel biocatalyst system for sulphide oxidation. treatment using an anaerobic biosystem. In Proceedings of
Biotechnology Progress 17: 439-446. the 3rd Canadian Organic Residuals and Biosolids
Mesa, M. M., M. Macías and D. Cantero. 2002. Biological iron Management Conference, 502-513. Calgary, AB.
oxidation by Acidithiobacillus ferrooxidans. Chemical and Sorokin, Y.D., P.T. Tourova, V.T. Kolganova, A.K. Sjollema
Biochemical Engineering Quarterly 16(2): 69-73. and G.J. Kuenen. 2002. Thioalkalispira microaerophila
Monteith, H., M. Béland and W. Parker. 2005. Assessment of gen. nov., sp. nov., a novel lithoautotrophic, sulfur-
economic viability of digester gas cogeneration at Canada’s oxidizing bacterium from a soda lake. International Journal
largest wastewater treatment plants. CD of Proceedings of of Systematic and Evolutionary Microbiology 52: 2175-
the 34th Water Environment Association of Ontario Annual 2182.
Conference. Huntsville, ON.
Sreekrishnan, R.T. and D.R. Tyagi. 1996. A comparative study
Morgan-Sagastume, M.J., A. Noyola, S. Revah and J.S. Ergas. of the cost of leaching out heavy metals from sewage
2003. Changes in physical properties of a compost biofilter sludges. Process Biochemistry 31(1): 31-41.
treating hydrogen sulfide. Journal of Air & Waste
Management Association 53: 1011-1021. Stanier, R.Y., J.L. Ingraham, M.L. Wheelis and P.R. Painter.
1986. The Microbial World, 5th edition. Englewood Cliffs,
MSDS 1996. Material Safety Data Sheet for Hydrogen Sulfide. NJ.: Prentice-Hall, Inc.
Murray Hill, NJ: BOC Gases. http://www.vngas.com/
pdf/g94.pdf (2006/02/20) Sublette, K.L. and N.D. Sylvester. 1987. Oxidation of hydrogen
sulfide by continuous cultures of Thiobacillus denitrificans.
Nishimura, S. and M. Yoda. 1997. Removal of hydrogen sulfide
Biotechnology and Bioengineering 29: 753-758.
from an anaerobic biogas using a bio-scrubber. Water
Science and Technology 36(6-7): 349-356. Syed, M.A. 2003. Effect of tube size and light source on
Odintsova, E.V., H.W. Jannasch, J.A. Mamone and T.A. performance of a fixed-film photobioreactor for conversion
Langworthy. 1996. Thermothrix azorensis sp. nov., an of hydrogen sulfide to elemental sulfur. Unpublished Ph.D.
obligately chemolithoautotrophic, sulfur-oxidizing, thesis. University of Windsor, Windsor, ON.
thermophilic bacterium. International Journal of Systematic Syed, M.A. and P.F. Henshaw. 2003. Effect of tube size on
Bacteriology 46(2): 422-8. performance of a fixed-film tubular bioreactor for
Oyarzún, P., F. Arancibia, C. Canales and E.G Aroca. 2003. conversion of hydrogen sulfide to elemental sulfur. Water
Biofiltration of high concentration of hydrogen sulphide Research 37(8): 1932-38.
using Thiobacillus thioparus. Process Biochemistry 39: 165- Syed, M.A. and P.F. Henshaw. 2005. Light emitting diodes and
170. an infrared bulb as light sources of a fixed-film tubular
Overmann, J. 2001. Diversity and ecology of phototrophic photobioreactor for conversion of hydrogen sulfide to
sulfur bacteria. Microbiology Today 28: 116-118. elemental sulfur. Journal of Chemical Technology and
Postgate, J.R. 1968. The sulphur cycle. In Inorganic Sulphur Biotechnology 80:119-123.
Chemistry, ed. G. Nickless, 259 - 279. Amsterdam, Takano, B., M. Koshida, Y. Fujiwara, K. Sugimori and S.
Netherlands: Elsevier Publishing Company. Takayangi. 1997. Influence of sulfur-oxidizing bacteria on
Prescott, M.L., P.J. Harley and A.D. Klein 2003. Microbiology, the budget of sulphate in Yugama crater lake, Kusatsu-
5th edition. New York, NY: McGraw-Hill Companies.. Shirane volcano, Japan. Biogeochemistry 38: 227-253.
Ross, C.C., T.J. Drake and J.L. Walsh. 1996. Handbook of Tchobanoglous, G., F.L. Burton and H.D. Stensel. 2003.
Biogas Utilization, 2nd edition. Florence, AL: SERBEP, c/o Wastewater Engineering. Treatment and Reuse, 4th edition.
General Bioenergy. McGraw-Hill Companies, New York, NY. Cap. 14: 1505-
Roth, S.H. 1993. Hydrogen sulfide. In Handbook of Hazardous 1532.
Materials. New York, NY: Academic Press. Vlasceanu, L., R. Popa and K.B. Kinkle. 1997. Characterization
Schieder, D., P. Quicker, R. Schneider, H. Winter, S. Prechtl of Thiobacillus thioparus LV43 and its distribution in a
and M. Faulstich. 2003. Microbiological removal of chemoautotrophically based groundwater ecosystem.
hydrogen sulfide from biogas by means of a separate Applied and Environmental Microbiology 63(8): 3123-
biofilter system: Experience with technical operation. Water 3127.
Science and Technology 48(4): 209-212. Van Niel, C.B. 1931. On the morphology and physiology of the
Sercu, B., D. Núñez, V.H. Langenhove, G. Aroca and W. purple and green sulfur bacteria. Archiv fur Mikrobiologie
Verstraete. 2005. Operational and microbiological aspects 3: 1-112.
of a bioaugmented two-stage biotrickling filter removing Wani, A.H., A.K. Lau and M.R. Branion. 1999. Biofiltration
hydrogen sulfide and dimethyl sulfide. Biotechnology and control of pulping odors – hydrogen sulphide: Performance,
Bioengineering 90(2): 259-269. microkinetics and coexistence effects of organo-sulfur
Sittig, M. 1978. Petroleum Refining Industry: Energy Saving species. Journal of Chemical Technology and
and Environmental Control. Park Ridge, NJ: Noyes Data Biotechnology 74: 9-16.
Corp.

2.14 LE GÉNIE DES BIOSYSTÈMES AU CANADA SYED et al.