Performance study of biofilter developed to treat H₂S from wastewater odour

Abstract

Biofiltration is an efficient biotechnological process used for waste gas abatement in various industrial processes. It offers low operating and capital costs and produces minimal secondary waste streams. The objective of this study was to evaluate the performance of a pilot scale biofilter in terms of pollutants’ removal efficiencies and the bacterial dynamics under different inlet concentrations of H₂S. The treatment of odourous pollutants by biofiltration was investigated at a municipal wastewater treatment plant (WWTP) (Charguia, Tunis, Tunisia). Sampling and analyses were conducted for 150 days. Inlet H₂S concentration recorded was between 200 and 1300 mg H₂S.m⁻³. Removal efficiencies reached 99% for the majority of the running time at an empty bed retention time (EBRT) of 60 s. Heterotrophic bacteria were found to be the dominant microorganisms in the biofilter. The bacteria were identified as the members of the genus Bacillus, Pseudomonas and xanthomonadacea bacterium. The polymerase chain reaction-single stranded conformation polymorphism (PCR-SSCP) method showed that bacterial community profiles changed with the H₂S inlet concentration. Our results indicated that the biofilter system, containing peat as the packing material, was proved able to remove H₂S from the WWTP odourous pollutants.

1. Introduction

Hydrogen sulphide is a colourless compound which can be found in natural gases as well as in volcanic gases and hot springs. Hydrogen sulphide is generated as a by-product by many industrial activities such as petroleum refining, pulp, paper manufacturing and wastewater treatment plant (Barona et al., 2004). It has a very typical smell of rotten eggs and it can be smelled at low concentrations (about 0.5 ppb) (Busca and Pistarino, 2003).

Several processes, such as scrubbing, adsorption and condensation, have been proposed for the treatment of waste gases. The physicochemical methods that have been used to remove pollutants from gas emissions have relatively high energy requirements and high chemical and disposal cost (YanLing et al., 2006).

Biofiltration is one of the most important biological processes of waste gas treatment and of odour control. Successful applications of this technology have been reported in wastewater treatment, petrochemical and tobacco industries (McNevin and Barford, 2000; Duan et al., 2006). This process has been increasingly regarded as the best available control technology owing to its high removal efficiencies, its low initial and operating cost and the easiness of its maintenance (Kikuchi, 2000; Busca and Pistarino, 2003). A biofilter consists of a container of organic material, populated with micro-organisms, through which contaminated air is usually passed upwards. The contaminated air stream is concurrently or counter-currently contacted with a liquid phase that provides nutrients and conditions to keep the viability and activity of the biofilm. In concurrent mode, contaminated air and liquid phase are sprayed from the top of the bed although this is kept to a minimum to prevent bed settling. But in the counter-current mode, contaminated air is sprayed from the bottom of the bed and the liquid phase is sprayed from the top of the bed. The contaminated gases pass through the biofilter bed, pollutants are transported into the biofilm where they are used by bacteria as a carbon source, an energy source or both (Ortiz et al., 2003; Ma et al., 2006). Through oxidative reactions, organic contaminants are converted to odourless compounds, such as carbon dioxide, water vapour, and organic biomass. When degrading inorganic compounds such as hydrogen sulphide, autotrophic bacteria utilise carbon dioxide as a carbon source resulting in the production of new biomass and sulphate or elemental sulphur (Barona et al., 2004; Andersson and Grennberg, 2001). Various bioactive media have been used as a packing bed such as compost, soil and horse manure. Packing medium should have a high surface area, high air and water permeabilities, and should provide a good surface of microbial growth. It plays an important role in air and water distribution, as well as the mass transfer (Song and Kinney, 2000). Biofilter performance is highly dependent on several parameters such as pH, temperature, adequate moisture content and inlet and outlet of H₂S. It also depends on the microbial community of the packed bed. The understanding of the microbial structure and diversity is important for biofiltration operation. However, only limited information is available on the microbial communities involved in the biofiltration of odourous pollutants. Currently, several molecular techniques, such as denaturing gradient gel electrophoresis (DGGE) and SSCP, have been used to study microbial communities. PCR-SSCP method offers a simple, inexpensive and sensitive method for studying the microbial diversity from several ecosystems. It can give a more objective picture of the bacterial community (Khelifi et al., 2009).

The objective of this study was to investigate the effects of the H₂S loading rate, firstly on the biofilter performance (e.g., elimination capacity, removal efficiency) and secondly on the bacterial diversity and dynamics. Various physicochemical and microbiological parameters were monitored, including H₂S concentration, moisture content, pH, Chemical oxygen demand (COD), sulphate concentration and colonies forming units (CFU). The bacterial communities of the biofilter were examined by the PCR-SSCP technique.

2. Material and methods

2.1. Experimental design

Biofilter system was designed according to previous work (Omri et al., 2011) and it is schematically displayed in Fig. 1. This Biofilter has a cubic form with a height of 2 m, an outside width of 5.4 m and an inside width of 4.9 m. It was packed with peat (German Company: NEVEMA GmbH) which has a height of 1 m and a porosity of 60%. The concentration of H₂S was unstable in the inlet of biofilter. It ranged between 200 and 1300 mg m⁻³. Therefore, the loading rate is sometimes higher than microbial capacity, affecting the microbial stability. These problems were countered to some extent by using higher EBRT (60 s). Treated wastewater was used as the nutrient solution and it was sprayed from a nozzle at the top of the biofilter. This irrigation maintains the bed moisture and provides the essential nutrients for the microbial activity, as well as the removal of the by-products produced during the biological reaction. The flow rate of the irrigation solution was constant.

Figure 1.

Pilot scale biofilter.

2.2. Odourous gas collection

The gas concentration on the sampling points in the biofilter system was periodically monitored. The gas samples were collected in Tedlar bags on the sampling points and were analysed immediately to avoid sample deterioration.

2.3. Sampling and analyses

Analyses were performed for the inlet and outlet gases, the irrigation liquid phase and the peat packed bed biofilter. Peat was divided into three layers (down, middle and up) for analyses.

2.4. Analytical methods and microbial load analyses

Inlet and outlet H₂S gas concentrations were continuously measured by titration using a standard potassium iodide–iodate titrant and a starch indicator (APHA, 1992). In addition, they were periodically measured by gas detector tubes (Kitagawa, Japan). The detection ranges of the H₂S gas tubes were 1–2000 and 1–200 ppm for Inlet and outlet H₂S gas concentrations, respectively.The removal efficiency was calculated using the following equation (Eq. (1)) (Omri et al., 2011):

$$\text{RE}=\frac{(C_{\text{in}}-C_{\text{out}})\times 100}{C_{\text{in}}}$$ (1)

where, RE: removal efficiency (%); C_in: inlet H₂S concentration (mg.m⁻³); C_out: outlet H₂S concentration (mg.m⁻³).

The pH was measured using a digital calibrated pH-metre (HANNA pH 210). The soluble COD was measured following standard methods (Khelifi et al., 2009). The moisture content of the compost-based biofilter was determined by taking a 1 g sample of the peat medium, then weighing and drying it for 24 h at 105 ± 0.5 °C.

As for microbial analyses peat samples (0.5 g) were taken from each of the three layers (down, middle and up) of the biofilter bed. They were mixed with 5 ml saline solution (0.9% w/v NaCl). The samples were then vortexed for 3 min. The cell numbers in the biofilter bed, expressed as colonies forming units per gram (CFU/g), were enumerated by traditional plate counting methods. Thiosulphate (TS) medium (composition in g L⁻¹: K₂HPO₄ 1.5; KH₂PO₄ 1.5; NH₄Cl 0.4; MgCl₂·7H₂O 0.8; CaCl₂ 2H₂O 0.1; Na₂S₂O₃·5H₂O 10) was used to enumerate the sulphur oxidising bacteria (SOB). The Plate Count Agar (PCA) and PDA (Potato Dextrose Agar) media were used for culturing heterotrophic bacteria and fungi, respectively. The obtained colonies on TS medium were characterised using morphology and biochemical tests according to Bergey’s Manual of Systematic Bacteriology (Kelly and Harrison, 1989). Biochemical tests used were: catalase, oxidase, urease, MR (methyl red), VP (Voges-Proskauer), glucose fermentation and H₂S production.

2.5. DNA extraction, and identification of the isolated colonies

The morphologically distinct bacterial isolates obtained by CFU counting were selected and purest clones were identified. DNA was extracted from cells by the heat shock method. The components necessary for PCR amplification with AmpliTaq DNA polymerase were added along with T7 and M13 primers (Khelifi et al., 2009). PCR products of the expected size were purified with the QIAquick PCR purification kit (Qiagen, Germany). Purified PCR products obtained from selected clones were sequenced.

2.6. SSCP analysis of the peat packed bed

The structure of the bacterial community of the peat packed bed biofilter was analysed by SSCP according to the protocol developed by Omri et al. (2011).

3. Results and discussion

Some parameters must be taken into consideration in order to understand their influence on the performance of the biofiltration system. The pH, moisture content, bacterial cell concentration (CFU.g⁻¹ dry peat), temperature, COD and H₂S concentration were thus measured weekly. These various measures were carried out on samples taken from gas, three layers of peat and liquid phase.

3.1. The performance of the biofiltration treatment

The feasibility of using a biofilter system to treat Hydrogen sulphide contaminated air from WTTP was investigated in this work. Inlet H₂S concentrations were varied from 200 to 1300 mg.m⁻³. Sampling and analyses were conducted for 150 days. The H₂S concentration was periodically determined in the inlet and outlet of the biofilter bed, then removal efficiency was calculated.

Results show that H₂S concentration decreased from a range of 200–1300 mg.m⁻³ (131–854 ppm) (inlet) up to a range of 5–120 mg.m⁻³ (3–78 ppm) (outlet), following the biofiltration treatment. As can be seen in Fig. 2(A), the average H₂S removal efficiencies of the biofilter were above 90% for the majority of the running time. For inlet H₂S concentrations ranging from 200 to 600 mg.m⁻³, higher removal efficiencies were still achieved at 99% and less than 5 mg.m⁻³ (5 ppm) of H₂S gas was detected in the outlet. H₂S removal efficiency declined up to 89% when the inlet H₂S concentrations ranged from 600–1300 mg.m⁻³.

Figure 2.

Global Performance of the biofilter bed treating H₂S containing gases during an operational period of 150 days: (A) (Removal efficiency () of H₂S containing gases in the biofilter Inlet (■) and outlet (♦) H₂S concentrations) and (B) (H₂S removal () and elimination capacity (EC) () as a function of the inlet H₂S loading in biofilter).

Results indicate that our biofilter system, containing peat as the filter material, removed H₂S with efficiencies ranged between 89% and 99% for H₂S inlet concentrations at the range of 200–1300 mg H₂S.m⁻³. In addition, outlet H₂S concentrations are very high. In an attempt to achieve the removal of H₂S from waste gas, a follow up treatment such as adsorption using activated carbon as a packing media may be required. This synthetic media has been the most extensively used material for physical adsorption of odour. The enhanced performance may manifest in higher removal efficiency with a decrease of outlet H₂S concentrations.

H₂S removal and elimination capacity as a function of the inlet H₂S loading was estimated and presented in Fig. 2(B). The elimination capacity is defined as the amount of pollutant degraded per unit of time, normalised to the volume of the packed bed. The inlet H₂S loading is defined as the amount of inlet gas per unit time and volume of packed bed (g.m⁻³.h⁻¹) (Rattanapana et al., 2009). We have found that H₂S removal was affected by the shock loading when the inlet H₂S loading suddenly increased from 11.7 to 60 g.m⁻³.h⁻¹. Moreover, the maximum elimination capacity was determined as 58 g H₂S.m⁻³.h⁻¹ in biofilter. On the other hand, when the biofilter had been fed with higher H₂S concentration, lower efficiencies were obtained. Exposure to high H₂S concentration could cause toxicity for microbial community of the peat packed bed, causing the decline of removal efficiency (Chaiprapat et al., 2011). Similarly, Kim et al. (2002) showed that the biofilter system with wood chips and granular activated carbon as a bed, were sensitive to H₂S shock loading. According to these authors, the removal efficiency decreased from 99% to 64% immediately when a H₂S loading rate increased from 6 to 10 g m⁻³.h⁻¹.

Several biofilter systems were previously studied. Representative cases are presented in Table 1.

Table 1.

Biological processes used for hydrogen sulphide removal.

Process type	Bed	Pollutant treated and inlet concentration	EBRTa	Removal efficiency (%)	Reference
Biofiltration	Wood chips, granular activated carbon	30–450 ppm H2S, 35–200 ppm NH3	20–60 s	75–99 (H2S) 30–92 (NH3)	Kim et al. (2002)
Biofiltration	Structured plastic packing	60–155 ppm H2S, 35–100 mg m−3 CS2	11 s	85–99 (H2S) 40–70 (CS2)	Cox and Deshusses (2002)
Biofiltration	Mature compost	100 ppmv H2S	50 s	90–100	Morgan-Sagastume and Noyola (2006)
Biofiltration	Activated sludge	5–100 H2S ppmv	2–21 s	94–99	Duan et al. (2006)
Biofiltration	Material made of pig manure and sawdust	0.03–0.32 g m−3 H2S	14–36 s	83–97	Barona et al. (2004)
Immobilised-cells bioreactor	Granular activated carbon inoculated with T. denitrificans	100 mg L−1 H2S	25 s	98	Ma et al. (2006)
Biofilter	Peat	200–1300 mg m−3 131(+/−0.01)−854(+/−0.08) ppmv H2S	60 s	94–99	This work

^aEBRT: empty bed retention time.

3.2. Change in pH, moisture and bacterial community of the peat-based biofilter

In this study, peat was used as the biofilter packing material to treat gases emitted from the (WWTP). The characterizations of the peat were based on the measurement of pH, moisture content and bacterial cell concentration (CFU.g⁻¹ dry peat). Samples were taken from three layers of packed bed (down, middle and up). The corresponding values are represented in Fig. 3 and Table 2.

Figure 3.

Time course of moisture (A) content and pH (B) in the three layers (down (■), middle () and upper (♦) of biofilter bed treating H₂S and the inlet H₂S concentration ().

Table 2.

Bacterial population at various positions on the peat bed of the biofilter at the start and the end of the experiment.

Position in biofilter	Cell concentration (CFU mL−1)
	2nd day (H2S inlet conc = 200 mg.m−3)		44th day (H2S inlet conc = 1300 mg.m−3)
	Sulphur oxidising bacteria	Heterotrophic bacteria	Sulphur oxidising bacteria	Heterotrophic bacteria
Up	6.04 105	7.26 107	1.6 104	7 106
Middle	9.4 106	12 108	2.24 104	23 107
Down	12 108	4.1 109	7.1 106	4 106

3.2.1. Moisture content

Moisture content of biofilter media is considered as a parameter necessary for its good performance (Morgan-Sagastume and Noyola, 2006; McNevin and Barford, 2000). Examination of Fig. 3(A) shows that the distribution of relative humidity at three layers of the biofilter was not uniform. It varied over time in each layer. Moisture content varied from 35% to 80% for the bottom layer, from 37% to 80% for the middle layer and from 33% to 80% for the upper layer. It was observed that humidification of the inlet gas was sufficient for maintaining moisture in the peat. It is worth noting that the moisture content is the most important parameter for biofilter viability. Optimal moisture content varied from 20% to 60%. If the moisture content of an organic material was too low, microbial activity decreased markedly. But at high moisture content, anaerobic zones appeared in the bed causing the decrease in the oxygen quantity necessary for biological activity. Consequently, the efficiency of peat bed to remove odour decreased noticeably when it became too wet (Cox and Deshusses, 2002; McNevin and Barford, 2000).

The obtained results indicated that during the deodourization process, our biofilter retained an average moisture content of 50%. As Chung (2007) notes that moisture content must be maintained between 40% and 60% in the biofilter to maintain biological activity, so, our results demonstrated that the peat packed bed has good water holding capacities. Similarly, Desshusses (1994) indicated that compost or peat, used as a support for biofiltration, has a large surface area, with a high potential retained moisture content and high sorption potential.

3.2.2. pH

pH also played a role in biofilters performance as can be seen in Fig. 2(B). The greatest spectrum of bacterial activity requires a near neutral pH. However, the sulphur oxidising bacteria preferred acidic pH 3.

According to Fig. 3(B), it was observed that the pH of the bottom and middle layers were remaining unstable over time. The changes in pH values of the bottom and middle layers were in the range of 4–7.8 and 3.3–7.5, respectively. However, for the upper layer, no significant change was observed. pH was at the range of 6–7.3. The declining of pH in the bottom and middle layers could be caused by the production of H⁺ and sulphate or sulphuric acid from the oxidation of H₂S (Eqs. (2) and (3)) (Kim et al., 2008).

$${\text{H}}_2\text{S}+2{\text{O}}_2\to \text{S}{\text{O}}_4^{2-}+2{\text{H}}^{+}$$ (2)

$${\text{H}}_2\text{S}+2{\text{O}}_2\to {\text{H}}_2\text{S}{\text{O}}_4$$ (3)

In the same frame, Chaiprapat et al., 2011 showed that the conversion of H₂S containing biogas to sulphuric acid through a complete oxidation by sulphur oxidising bacteria (SOB) caused the decrease of pH. They noticed that lowest pH provides favourable conditions for sulphide oxidisation. Our Results also showed that the biodegradation of H₂S gas was realised in the bottom and middle layers.

Besides, our data showed that lowest pH of the peat was found when the biofilter was fed with higher concentration of H₂S. As described in paragraph (3.1), removal efficiency dropped with increasing the inlet concentration of H₂S. So, it was found that the removal efficiency declining with acidification of the media is caused by the accumulation of sulphate and acidic products.

3.2.3. Microbial consortia on the peat packed bed

In this work, only indigenous microorganisms to the bed medium were used throughout the whole process. After cultivation in TS and PCA plates, all morphologically different colonies were picked from each plate. The quantitative estimates of the microbial population at various positions of each biofilter are shown in Table 2. A significant difference was found for the cell concentration of SOB between the three layers. Denser SOB community occurred in the bottom and the middle layers of the bed with cell concentration values of 12 10⁸ and 9.4 10⁶ CFU.g⁻¹ dry peat, respectively. However, SOB cell concentration in the upper layer (6.04 10⁵ CFU.g⁻¹ dry peat) was less than those in the two other layers. Similarly, high heterotrophic bacteria-concentration was observed in the bottom and the middle layers of the bed with cell concentration values of 4.1 10⁹ and 12 10⁸ CFU.g⁻¹ dry peat, respectively, while the heterotrophic bacteria concentration in the upper layer (7.26 10⁷ CFU.g⁻¹ dry peat) was the lowest among the three layers. It was also found that there are significant differences in the cell concentrations of SOB and heterotrophic between layers, independent of the H₂S concentration. These results reflect the H₂S inhibitory effect. In fact, for the different layers, cell concentration values of SOB and heterotrophic bacteria, decreased when H₂S inlet concentration increased. When the H₂S inlet concentration increased from 200 (day 2) to 1300 mg.m⁻³, cell concentration values of SOB decreased from 12 10⁸ to 7.1 10⁶ (CFU.g⁻¹ dry peat) in the bottom layer. It could be also explained that a pH drop provides favourable conditions for sulphur oxidation effect on bacterial growth. So, we can conclude that biological oxidation of H₂S was found in the middle and bottom layers. It is due to the fact that peat at the bottom layer of the bed received the most concentrated H₂S, as substrate from waste gas and nutrients from the liquid phase (treated wastewater). These conditions were favourable for the growth of heterotrophic bacteria. However, at the upper layer of the peat, substrate and nutrients were limited.

Based on morphological characteristics, only Gram negative bacteria were found on TS plates, while some Gram positive bacteria were found in PCA plates. In the TS plates, the three layers of the peat had different bacterial strains. Twelve isolates with distinct colony morphology were obtained: seven bacterial strains from the bottom layer, four bacterial strains from the middle layer and one isolate from the upper layer. Each isolate consisted of gram negative and small rod-shaped cells. It was observed that most of the isolated strains produce pigments on nutrient agar plates, with predominance of yellow pigmented bacteria. All isolated strains were oxidase (−) and catalase (+) except for two strains obtained from the bottom layer, which are oxidase (+) and catalase (+). Also, it was noticed that glucose was not fermented and there was no H₂S gas production. The obtained results indicated that the isolated strains were recognised as SOB because of their ability to use thiosulphate as a sole source of energy and then to be cultivated in TS medium.

Colonies with distinct morphologies from each medium (TS plates and PCA plates) were subjected to 16S rRNA sequencing for identification. The closest relative was identified by comparison with the Genebank database (Table 3). The identification of distinct pure colonies showed that bacterial colonies of the peat packed bed were members of the Bacillus and Pseudomonas genus such as Bacillus sp., Xanthomonadaceae bacterium and Pseudomonas putida. These bacteria were common bacterial species which were known to be involved in the sulphur odour treatment system (Xie et al., 2009; YanLing et al., 2006).

Table 3.

Analysis of the 16S rRNA gene sequence from the isolated colonies.

Family	Closest	Relatives similarity (%)	Accession number	Sources
Proteobacteria	Acinetobacter junii	99	AB101444	Activated sludge
Bacillacea	Bacillus sp.	99	JF734317	Soil
Xanthomonadaceae	Xanthomonadacea bacterium	99	FJ985737	Soil
Pseudomonadaceae	Pseudomonas putida	98	AB512773.1	Wastewater, soil

3.3. Bacterial dynamics in the biofilter system

In order to study the change of the bacterial population structure in the biofilter bed, PCR-SSCP analyses were realised for peat samples taken when the system was fed by the highest and the lowest H₂S concentrations. Profiles of the bacterial communities are shown in Fig. 4. Results showed that bacterial community profiles changed simultaneously with the increase of the H₂S loading rates. SSCP profiles reveal the high bacterial diversity in the peat packed bed.

Figure 4.

Dynamics of single strand conformation polymorphism (SSCP) patterns of bacterial 16S rDNA region amplification products of the biofilter bed treating H₂S containing gases: (A) at lowest H₂S concentration (2nd) and (B) at highest H₂S concentration (44th).

The SSCP pattern (A) is a picture of the bacterial community which had been present at the lowest H₂S concentration (200 mg.m⁻³ for the 2nd day). It revealed that the diversity of the ecosystem is large, with at least 30 distinguishable peaks. 26 peaks of this SSCP pattern co-migrated with peaks of SSCP pattern (B) obtained after increasing H₂S concentration (1300 mg.m⁻³ for the 44th day). Some peaks noticeably increased and became prominent. However, other peaks disappeared or noticeably decreased. Increasing H₂S loading caused a significant change of the SSCP pattern (B) with a strong decrease of the microbial diversity. The persisted peaks represented bacteria which were able to tolerate and to grow under high H₂S loading these conditions.

The reciprocal Simpsons diversity index was calculated based on the PCR-SSCP results. The value ranged between 8.5 and 13.4 (which offer towards 1 as the lowest possible value) for the different samples. At the lowest H₂S concentration (200 mg.m⁻³ for the 2nd day), the peat packed bed had the highest diversity (1/D = 13.4). The diversity index value of the peat at the highest H₂S concentration (1300 mg.m⁻³ for the 44th day) decreased to 8.5. This result indicated that some species of bacteria disappeared and fewer bacteria survived.

The evenness values of the different samples were 0.55 and 0.3 for peat at the lowest and the highest concentrations of H₂S (which offers towards 0), respectively. The obtained result indicated that few microorganisms dominated the bacterial populations in the peat packed bed at different H₂S concentrations.

This result showed a positive correlation between the diversity and the reactor performances. With the increase of H₂S concentrations, bacterial community profiles simultaneously changed. These changes in bacterial diversity affected the performances of the reactor in terms of the change of bacteria dominance and diversity from pattern (A) to pattern (B). In fact, H₂S removal efficiency decreased from 98% to 89%, when the biofilter was fed by H₂S inlet concentration from 200 to 1300 mg.m⁻³, respectively. These decreases of reactor performances were, first due to inhibitory effects of high concentration of the H₂S or intermediate products (sulphate) on bacteria. Second, they may be explained by the changes of diversity that caused the shift of a number of bacteria and the decrease of biological activity. In fact, the increases of the H₂S concentration induced a decrease of the diversity and allowed the growth of few bacterial species. These changes caused the selection of a few bacterial populations, in relation with the toxicity of the accumulated H₂S and sulphate products on microorganisms. A simultaneous study of the biofilter performance and the change in bacterial diversity has been used to propose a correlation between a specific activity of the process and bacterial community.

The physicochemical results were correlated with microbiological results to explain the biofilter performances. In fact, pH and CFU counting determined from the biofilter lead to the same results as shown in Fig. 3(B), Table 2 and the changes of the bacterial diversity and dominance presented in the different SSCP patterns at the different H₂S loading (Fig. 4).

3.4. Characterisation of liquid phase of biofilter treating H₂S

A liquid phase may maintain adequate moisture content, dilute toxic metabolic products, provide additional nutrients for biological growth and provide buffering for pH stability. Dilution has been found to be particularly important for hydrogen sulphide biofiltration. In this study, peat medium was irrigated by treated wastewater of the WWTP. Characterisations of liquid phase are based on the measurement of pH, COD and sulphate concentration of the percolate. The corresponding values are depicted in Fig. 5.

Figure 5.

Time course of pH (♦) and sulphate concentration (■) (A) and COD (B) (inlet () and outlet (□)) in the percolate of the biofilter treating H₂S containing gases during an operational period of 150 days.

According to Fig. 5(A), pH of liquid phase dropped from 7 to 1.5 during running time. During the later period of the experiment, the pH of the medium was quite acid because of the accumulation of the large amount of sulphuric acid. Hence, sulphate, as a product of microbial oxidation of H₂S, could be responsible for the pH decrease.

As can be seen in Fig. 5(A), sulphate concentration in the percolate varied with pH values. When pH of the liquid varied from 3 to 1.5, sulphate concentration increased from 760 to 1400 mg L⁻¹. Hence, a correlation between pH and sulphate concentration of the liquid could be noted. On the other hand, the decline of pH was due to the production of H⁺ and sulphate as primary product, and sulphuric acid as a second product, from the oxidation of H₂S. So, only if sulphur concentration is relatively high, a relevant acidification of the biofilter is found. Similarly, Chaiprapat et al. (2011) observed the decrease of pH of the liquid phase with the decrease of removal efficiency. This result could be caused by the lower solubility of the gases in the liquid film when it became more acidic.

Fig. 5(B) shows the monitoring of COD concentration in the inlet and outlet of the liquid phase during running time. COD of the water at the inlet of the bed was in the range of 20–30 mg L⁻¹. The COD of the liquid increased at the outlet of the bed, and it was in the range of 350–680 mg L⁻¹, while the Tunisian standards of wastewater allow only COD of 90 mg L⁻¹ (Tunisian Norms, 1989b). Therefore, the percolate must be treated in order to be rejected without public menace or used in irrigation.

4. Conclusion

Our results demonstrated that a biofilter system, containing peat as the packing material, was able to remove H₂S from the WWTP odourous pollutants. The removal efficiency of H₂S reached up to 99% at an empty bed retention time (EBRT) of 60 s in pilot scale. pH and inlet H₂S concentration affect the performance of the biofilter. It was found that the peat was suitable for the treatment of H₂S. It provides nutrient-rich environment for bacterial growth because of its good water holding capacity and buffering effect. The most abundant microorganism in the biofilter bed was found to be the heterotrophic bacteria such as Pseudomonas and xanthomonadacea bacteruim. The molecular fingerprint performed in this study revealed the high bacterial diversity of the reactor at different H₂S concentrations. It has been shown that bacterial community profiles changed simultaneously by increasing H₂S concentrations. These changes affected the biofilter performances.