ABSTRACT
Anaerobic treatment of sulfur-rich wastewater is challenging because sulfide greatly inhibits the activity of anaerobic microorganisms, especially methanogenic archaea. We developed an internal phase-separated reactor (IPSR) that removed sulfide prior to methanogenesis by gas stripping using biogas produced in the reactor. The IPSR was fed with synthetic wastewater containing a very high sulfide concentration of up to 6,000 mg S L−1 with a chemical oxygen demand (COD) of 30,000 mg L−1. The IPSR was operated at an organic loading rate of 5–12 kg COD m−3 day−1 at 35 °C. The results show that the sulfide concentration was reduced from 6,000 mg S L−1 in the influent to <700 mg S L−1 in the first-stage effluent. The second-stage effluent contained <400 mg S L−1. As a result of effective sulfide removal by its gas stripping function, the IPSR had a COD removal efficiency of >90% over the entire experimental period. High-throughput 16S rRNA gene sequencing revealed that the major anaerobic archaea were Methanobacterium and Methanosaeta, which are frequently found in high-rate anaerobic reactors. Thus, the IPSR maintains these microorganisms and achieves high-process performance even when fed wastewater with very high sulfide concentrations.
HIGHLIGHTS
An internal phase-separated reactor (IPSR) was developed.
The IPSR was used to treat sulfide-rich wastewater (3,000–6,000 mg S L−1).
Sulfide was removed by gas stripping using biogas produced in the IPSR.
The IPSR was operated with 5 kg chemical oxygen demand m−3 day−1 and 6,000 mg S L−1.
The IPSR maintained highly efficient organisms even under high sulfide conditions.
INTRODUCTION
Sulfur-containing wastewater is produced by the pharmaceutical industry (Li et al. 2015), petroleum refineries (Altaş & Büyükgüngör 2008), paper mills (Janssen et al. 2009), paper recycling plants (Damianovic et al. 2018), slaughterhouses (Yan et al. 2018a), vinasse production (Barrera et al. 2014), and other processes (Pol et al. 1998; Robles et al. 2020; Shao et al. 2024). Research on the treatment of wastewater with a high sulfur concentration has increased over the past three decades and has focused on biological methods (Ding & Zeng 2022). However, the implementation of anaerobic reactors for treatment of sulfur-rich wastewater is challenging because sulfide (HS−) produced by sulfate () is an inhibitor of various microorganisms, including sulfate-reducing bacteria and methane-producing archaea (Harada et al. 1994; O'Flaherty et al. 1998; Pol et al. 1998; Chen et al. 2008; Lens et al. 2010; Hao et al. 2014). The level of sulfide inhibition depends on the sulfide concentration and other environmental conditions, such as the pH during anaerobic digestion (Koster et al. 1986; Omil et al. 1995; Chen et al. 2008; Lee & Hwang 2019).
Various technologies have been used to remove sulfur from wastewater before or during the anaerobic digestion process. These technologies can be categorized as physicochemical techniques (e.g., gas stripping), chemical reactions (e.g., precipitation), and biological conversion (Chen et al. 2008; Robles et al. 2020). However, these sulfur removal processes need additional equipment (e.g., gas-stripping devices) and/or materials (e.g., chemicals). For successful application, the treatment technologies should be environmentally friendly, simple, easy to manage, and inexpensive.
To fulfill the need for appropriate technology, we recently developed a new reactor configuration that we call an internal phase-separated reactor (IPSR) (Onodera et al. 2022). The IPSR removes H2S by gas stripping using biogas produced during anaerobic treatment (Onodera et al. 2022). Removal of H2S from wastewater by gas stripping is possible because the biogas has a relatively low H2S concentration compared with the wastewater. This function is attributed to the structure of the IPSR. The IPSR consists of a first stage for gas stripping in an upper section of the reactor and a second stage for methane production in a lower section. A gas–liquid partitioning (GLP) valve is installed between the stages as a phase separator. The GLP valve stops liquid from flowing between the first and second stages and supplies biogas from the second stage to the first stage.
Previous experiments using a laboratory-scale reactor verified that the IPSR showed superior performance for sulfide removal compared with a control reactor when the sulfate concentration was between 400 and 800 mg S L−1 (Onodera et al. 2022) and the sulfide concentration was between 400 and 600 mg S L−1 (Onodera et al. 2023). After using the IPSR for sulfide removal, the sulfide concentration in the final effluent from the IPSR was half to one-third that obtained with the control reactor (Onodera et al. 2023). However, the various types of industrial wastewater contain different concentrations of sulfides (Shao et al. 2024). After spent caustic, the type of wastewater with the highest sulfide concentrations is distillery spent wash, which contains 102–130 g chemical oxygen demand (COD) L−1 and 1,100 to 3,000 mg SO4 L−1 (360 to 1,000 mg S L−1) (Karhadkar et al. 1990). For stillage from ethanol production, the average COD and total sulfur concentration as SO4 for various feedstocks are 91.1 g COD L−1 and 3,716 mg SO4 L−1 for beet molasses, 30.4 g COD L−1 and 1,356 mg SO4 L−1 for cane juice, 84.9 g COD L−1 and 3,478 mg SO4 L−1 for cane molasses, and 61.3 g COD L−1 and 651 mg SO4 L−1 for cellulosic feedstocks (Wilkie et al. 2000). Thus, for successful application of the IPSR to various types of sulfur-containing wastewater, it is necessary to evaluate the process performance when treating wastewater with much higher concentrations of sulfide than in our previous studies.
In this study, the process performance of IPSR for the treatment of wastewater containing much higher concentrations of sulfides (3,000–6,000 mg S L−1) than in our previous study (Onodera et al. 2023). The process performance was evaluated for sulfide removal, biogas production, and H2S removal by gas stripping. The prokaryotic microbial community under high sulfide concentrations was also determined.
METHODS
Synthetic wastewater
Synthetic sulfide-rich wastewater was prepared with a fixed COD of 30,000 mg L−1. The organic compounds were 12,000 mg L−1 sucrose, 9,000 mg L−1 CH3COONa, 4,500 mL L−1 propanoic acid, and 3,000 mg L−1 peptone. Sodium sulfide (Na2S·9H2O) was used as a sulfide source (inhibitor) at concentrations between 0 and 6,000 mg S L−1. The following compounds were added to the wastewater: 320 mg L−1 Na2SO4, 2,000 mg L−1 NaHCO3, 150 mg L−1 KH2PO4, 150 mg L−1 CaCl2·2H2O, 400 mg L−1 MgCl2·6H2O, 300 mg L−1 KCl, 110 mg/L NH4Cl, 3.93 mg L−1 FeCl2·4H2O, 0.17 mg L−1 CoCl2·6H2O, 0.07 mg L−1 ZnCl2, 0.06 mg L−1 H3BO3, 0.5 mg L−1 MnCl2·4H2O, 0.04 mg L−1 NiCl2·6H2O, 0.03 mg L−1 CuCl2·2H2O, and 0.03 mg L−1 NaMoO4·2H2O. An anti-foaming reagent (Antifoam SI, FUJIFILM Wako Pure Chemical Corp., Osaka, Japan) was also added. The wastewater composition was established in our previous study (Onodera et al. 2023). The synthetic wastewater was prepared freshly each week and kept at ambient temperature.
Reactor setup and process operation
Operating conditions of the internal phase-separated reactor during the experiment
. | . | First period . | Second period . | Third period . | Fourth period . | Fifth period . |
---|---|---|---|---|---|---|
Operation period | Day | 7 | 7 | 7 | 7 | 14 |
Flow rate | L day−1 | 2.06 | 2.06 | 1.54 | 1.03 | 0.85 |
Influent concentration | ||||||
COD | mg L−1 | 30,000 | 30,000 | 30,000 | 30,000 | 30,000 |
Sulfate | mg S L−1 | 72 | 72 | 72 | 72 | 72 |
Sulfide | mg S L−1 | 0 | 3,000 | 4,000 | 5,000 | 6,000 |
Second stage | ||||||
HRT | h | 60 | 60 | 80 | 120 | 145 |
COD loading | g COD L−1 day−1 | 12.0 | 12.0 | 9.0 | 6.0 | 5.0 |
. | . | First period . | Second period . | Third period . | Fourth period . | Fifth period . |
---|---|---|---|---|---|---|
Operation period | Day | 7 | 7 | 7 | 7 | 14 |
Flow rate | L day−1 | 2.06 | 2.06 | 1.54 | 1.03 | 0.85 |
Influent concentration | ||||||
COD | mg L−1 | 30,000 | 30,000 | 30,000 | 30,000 | 30,000 |
Sulfate | mg S L−1 | 72 | 72 | 72 | 72 | 72 |
Sulfide | mg S L−1 | 0 | 3,000 | 4,000 | 5,000 | 6,000 |
Second stage | ||||||
HRT | h | 60 | 60 | 80 | 120 | 145 |
COD loading | g COD L−1 day−1 | 12.0 | 12.0 | 9.0 | 6.0 | 5.0 |
Schematic diagram of the internal phase-separated reactor (IPSR): (1) first stage (upper section); (2) second stage (lower section); (3) phase separator (gas–liquid partitioning valve); (4) water jacket; (5) water seal; (6) H2S trapping pellet; (7) gas meter; (8) influent tank; (9) pump; and (10) gas sampling port.
Schematic diagram of the internal phase-separated reactor (IPSR): (1) first stage (upper section); (2) second stage (lower section); (3) phase separator (gas–liquid partitioning valve); (4) water jacket; (5) water seal; (6) H2S trapping pellet; (7) gas meter; (8) influent tank; (9) pump; and (10) gas sampling port.
The first stage, which was used for H2S gas stripping, was filled with synthetic wastewater without sludge before start-up. The second stage, which was used for methane production, was filled with mesophilic granular sludge derived from an upflow anaerobic sludge blanket (UASB) reactor operated at an OLR of 15 kg COD m−3 day−1 under mesophilic conditions (35 °C) to eliminate the start-up period. The UASB was fed with synthetic wastewater that had a similar composition to the wastewater used in the present study. The operating conditions and reactor performance are detailed in Supporting Information (Table S1 and Figure S1). The sludge was transferred to the reactor, and the operation was started within 30 min. The sludge contained 69.2 g L−1 total suspended solids (TSS) and 63.7 g L−1 volatile suspended solids (VSS) (VSS/TSS = 0.92). The total volume of the sludge was 4.0 L (277 g of TSS and 255 g of VSS). After the experiment, the sludge mass in the IPSR contained 248 g of TSS and 225 g of VSS.
The effluent was recirculated to the second-stage influent at twice the flow rate (recirculation rate: 200%). A pH sensor was placed in the upper portion of the first section to maintain the pH at <5.9 by feeding HCl solution (3 mol L−1) into the influent line. The operation was divided into five periods according to the sulfide concentration (0–6,000 mg L−1) and OLR (Table 1). The influent COD concentration was fixed, and the flow rate was changed to adjust the HRT and OLR. The operation was conducted for 1 week for each period, except for the fifth period, which was 2 weeks. Between each operation period from one to five, some of the wastewater in the first stage was replaced with new wastewater with a high sulfide concentration to ensure the desired high sulfide concentration in the second stage was reached quickly.
Chemical analysis
Liquid samples of the influent, first-stage effluent, and second-stage effluent were collected. The pH was measured immediately after sampling using a pH meter (9625-10D, D-74, LAQUA, Kyoto, Japan). The dichromate COD and sulfate (SO4) were analyzed using a Hach water quality analyzer (DR1900, Hach, Loveland, CO, USA). The COD was measured using Hach testing kits (TNT822 and TNT823, Hach) after sulfide removal by sulfuric acid addition, followed by purging with N2 gas. The SO4 concentration was analyzed using a SulfaVer 4 reagent powder pillow (Hach). Sulfide in solution was determined using a total sulfide detector tube (coefficient of variation: 5%; 201H, HYDROTEC-S 330, GASTEC, Kanagawa, Japan). To prevent volatilization of H2S, liquid samples were taken directly from the tubes using a syringe while preventing contact with air and then stored in the sealed syringes until required for analysis. The TSS and VSS were determined using standard methods (APHA 2005).
Microbial community analysis
Samples of 0.5 mL of seeding sludge (before the experiment) and retained sludge (on day 42 after the initiation of the experiment) were collected for microbial community analysis. The collected sludge samples were stored at −80 °C until required for DNA extraction and subsequent 16S rRNA gene sequencing. The DNA extraction and 16S rRNA gene sequencing were performed at the Bioengineering Lab. Co., Ltd (Kanagawa, Japan). Briefly, the collected sludge samples were initially freeze-dried (VD-250R freeze dryer, TAITEC Corporation, Saitama, Japan) and then mixed vigorously at 1,500 rpm for 2 min using a multi-beads shocker (Yasui Instruments, Osaka, Japan). Next, Lysis Solution F (a mixture of sodium dodecyl sulfate and EDTA-2Na; Nippon Gene, Tokyo, Japan) was added to each sample. The samples were incubated at 65 °C for 10 min for cell lysis and then centrifuged at 12,000 ×g for 2 min. The supernatant was collected. DNA was purified from the supernatant using a Lab-Aid824s DNA extraction kit (Zeesan, Xiamen, China). The amplicon libraries of the V4 region of prokaryotic 16S rRNA genes were prepared by a two-step-tailed polymerase chain reaction method with 515F/806R primers (Caporaso et al. 2010). The resulting libraries were sequenced using the MiSeq platform with MiSeq reagent kit v3 (Illumina, San Diego, CA, USA). The obtained reads were trimmed with FASTX-Toolkit (ver. 0.0.14) and the amplicon sequence variants were generated using DADA2 implemented in QIIME 2 (ver. 2022.8) (Bolyen et al. 2019). The taxonomic classification of each amplicon sequence variant was performed using the SILVA 138 database (Quast et al. 2013). The obtained raw 16S rRNA gene amplicon sequence data are available in the DNA Data Bank of Japan Sequence Read Archive (https://ddbj.nig.ac.jp) under accession numbers DRR527022–DRR527023.
RESULTS AND DISCUSSION
Process performance
Evolution of the influent sulfide concentration (a), organic loading rate (OLR) in the second stage of the IPSR (b), pH (c), COD (d), sulfate in first- and second-stage effluents (e), and sulfide in first- and second-stage effluents (f).
Evolution of the influent sulfide concentration (a), organic loading rate (OLR) in the second stage of the IPSR (b), pH (c), COD (d), sulfate in first- and second-stage effluents (e), and sulfide in first- and second-stage effluents (f).
Although the sulfide concentration in the second-stage effluent in the present study was <400 mg S L−1, the concentration of free H2S was lower (39 mg S L−1) at pH 7.8 because the H2S/HS− ratio was low. Free H2S is a stronger inhibitor than HS−. The maximum free H2S concentration recommended to avoid severe inhibition is 125 mg S L−1 (Omil et al. 1995). In another study, the free H2S concentrations leading to 50% inhibition were 250 mg S L−1 at pH 6.4–7.2 and 90 mg S L−1 at pH 7.8–8.0 (Koster et al. 1986). Therefore, the pH in the IPSR could alleviate H2S inhibition even though the sulfide concentration was 400 mg S L−1 in the second stage of the IPSR.
Biogas production and H2S removal
Evolution of biogas production after H2S removal (a), CH4 concentration after H2S removal (b), and CH4 production as COD based on the wastewater volume (c).
Evolution of biogas production after H2S removal (a), CH4 concentration after H2S removal (b), and CH4 production as COD based on the wastewater volume (c).
Evolution of the H2S concentration in the biogas (a) and H2S removed from the IPSR using biogas (b).
Evolution of the H2S concentration in the biogas (a) and H2S removed from the IPSR using biogas (b).
The H2S removal efficiency using biogas was calculated from the biogas production and H2S content (Figure 4(b)). During the experiment, the H2S removal efficiency increased with increases in the influent sulfide concentration. The IPSR provided better H2S removal when the influent sulfide concentration was high rather than low. In the IPSR, the use of biogas is a key feature, and the amount of biogas is an important factor affecting the gas stripping efficiency. The sulfide removal mechanism involves (1) the production of biogas with organic matter removal, (2) the removal of H2S by gas stripping using the biogas, and (3) the removal of organic matter under low sulfide conditions (Onodera et al. 2022).
Sulfur mass balance

Sulfur mass balance in the IPSR in the second period (a), third period (b), fourth period (c), and fifth period (d).
Sulfur mass balance in the IPSR in the second period (a), third period (b), fourth period (c), and fifth period (d).
Prokaryotic microbial community
Microbial composition before (left) and after (right) the experiments (>1% relative abundance on the genus level). Phylum names are shown in the middle of the plot.
Microbial composition before (left) and after (right) the experiments (>1% relative abundance on the genus level). Phylum names are shown in the middle of the plot.
Applicability of the IPSR for wastewater with high sulfide concentrations
The laboratory-scale experiments showed that the IPSR removed H2S without needing additional equipment for gas striping. This is an important feature of the IPSR compared with other technologies. In the conventional two-stage process, acidification and sulfate reduction are performed in the first stage and methanogenesis in the second stage, and sulfide removal can be performed before methanogenesis to reduce inhibition (Elferink et al. 1994; Hao et al. 2014). For instance, sulfide removal by gas striping has been achieved using a UASB reactor equipped with a sulfide-striping device (Yamaguchi et al. 1999). For gas striping, an additional operation, such as biogas or nitrogen purging, is required (Lens et al. 2010). A drawback of the two-stage process is the energy consumption of mechanical equipment for H2S stripping (Hao et al. 2014). By contrast, the IPSR requires only a pH control agent for the first stage, and this pH control agent is generally not needed when the pH of the influent wastewater is low.
A potential limitation of the IPSR is that sulfide removal depends on the quantity of biogas generated from the degradation of organic matter in wastewater. Thus, the sulfide concentration and sulfide/COD ratio need to be considered for the application of the IPSR. Sulfur-containing wastewater often has a high COD, for example, distillery wastewater has a COD of 38–56 g L−1 and a sulfate concentration of 1,100 to 5,000 mg S L−1 (Barrera et al. 2014). In the present study, the IPSR was applied to wastewater with a higher sulfide concentration (6,000 mg S L−1) and lower COD (30 g L−1) than distillery wastewater. Our results showed that the performance of the IPSR was sufficient, even with the high sulfide concentration.
This experiment showed that the IPSR achieved high COD removal efficiency. This high performance was attributed to the optimum pH conditions for H2S removal in the first stage and methane production in the second stage. The optimum pH was maintained because the GLP valve prevented the mixing of the liquid between the first and second stages. The simple structure of the GLP valve will allow for the construction of large IPSRs and full-scale reactors.
Although this experiment was conducted over a short period, long-term stability is important for anaerobic treatment technologies. The primary function of the IPSR is gas stripping, which will occur if biogas is produced. Therefore, long-term stability of the IPSR can be expected even under high sulfide concentrations because of the reduction of H2S by gas stripping. For real wastewater treatment, setting an appropriate OLR according to sulfide concentration and /COD ratio will ensure that the performance of the IPSR is stable over the long term.
CONCLUSIONS
The IPSR achieved high COD removal efficiency of >90% at an OLR of 5 to 12 kg COD m−3 day−1. Gas stripping was effectively performed and resulted in a low sulfide concentration of 700 mg S L−1 in the first-stage effluent and 400 mg S L−1 in the second-stage effluent. The maintenance of a prokaryotic microbial community, similar to that of high-rate anaerobic reactors, was observed even at high sulfide concentrations. Overall, this study shows that the IPSR can achieve high COD removal efficiency for wastewater with sulfide concentrations as high as 6,000 mg/L by its gas stripping function using biogas produced in the reactor.
ACKNOWLEDGEMENTS
We thank Mrs Naoko Okubo for her support with the experiments. We thank Gabrielle David, PhD, from Edanz (https://jp.edanz.com/ac) for editing a draft of this manuscript.
FUNDING
This study was supported by KAKENHI [grant number JP18K04416] from JSPS and the Steel Foundation for Environmental Protection Technology, Japan.
DATA AVAILABILITY STATEMENT
All relevant data are included in the paper or its Supplementary Information.
CONFLICT OF INTEREST
T.O. and K.S. have patent #Japanese Patent No. 6029081 issued to Licensee.