{"id":8251,"date":"2026-04-03T12:57:52","date_gmt":"2026-04-03T12:57:52","guid":{"rendered":"https:\/\/www.proefschriftmaken.nl\/portfolio\/karine-kiragosyan\/"},"modified":"2026-04-23T08:56:51","modified_gmt":"2026-04-23T08:56:51","slug":"karine-kiragosyan","status":"publish","type":"us_portfolio","link":"https:\/\/www.proefschriftmaken.nl\/en\/portfolio\/karine-kiragosyan\/","title":{"rendered":"Karine Kiragosyan"},"content":{"rendered":"","protected":false},"excerpt":{"rendered":"","protected":false},"author":8,"featured_media":14030,"comment_status":"closed","ping_status":"closed","template":"","meta":{"_acf_changed":false,"footnotes":""},"us_portfolio_category":[45],"class_list":["post-8251","us_portfolio","type-us_portfolio","status-publish","has-post-thumbnail","hentry","us_portfolio_category-new-template"],"acf":{"naam_van_het_proefschift":"Maximization of sulfur formation in the presence of organic sulfur compounds in a dual bioreactor gas desulfurization system","samenvatting":"Dit proefschrift richt zich op het optimaliseren van biologische gasontzwaveling, in het bijzonder voor gasstromen die waterstofsulfide (H2S) en organische zwavelverbindingen (zoals thiolen) bevatten. De belangrijkste doelstelling was het maximaliseren van de vorming van elementaire zwavel in een duaal bioreactorsysteem genaamd Thiopaq Ultra.\n\nVier belangrijke resultaten werden bereikt:\n1. Er is een methodologie ontwikkeld voor het selecteren van de juiste biomassa voor het opstarten van installaties. Hierbij werd de parameter \u03b1 ge\u00efntroduceerd, die de verhouding tussen de enzymen FCC en SQR beschrijft. Een lage \u03b1-waarde bleek een voorspeller voor een hoge zwavelproductie.\n2. Het nieuwe duaal bioreactorsysteem (met een extra anaerobe bioreactor) werd ge\u00ebvalueerd. Dit systeem bleek stabieler en effectiever in het verwerken van schommelingen in de H2S-toevoer en de aanwezigheid van toxische thiolen. In de duale opstelling steeg de zwavelselectiviteit naar 90 mol% bij hoge thiolbelastingen, vergeleken met 75% in de traditionele opstelling.\n3. De toevoeging van dimethyldisulfide (DMDS) bleek sulfaatvorming effectief te remmen. Door de biomassaconcentratie te verhogen, kon de vorming van thiosulfaat worden geminimaliseerd, wat resulteerde in een zwavelselectiviteit van circa 96 mol%.\n4. Er werd een nieuwe 'feedforward' control-strategie voor de zuurstoftoevoer ontwikkeld op basis van een vaste O2\/H2S-verhouding van circa 0,60. Deze strategie is minder gevoelig voor verstoringen door thiolen dan de traditionele ORP-meting en leidde tot een stabiele zwavelselectiviteit van meer dan 95 mol%, zelfs bij vari\u00ebrende gasstromen.","summary":"7.1. Introduction\n\nThe growth of the global population and its associated increased demands for energy, food, and water has resulted in the intensification of industry and land use and hence loss of biodiversity and climate change. The overuse of natural resources, anthropogenic gas emissions and wastewater discharges into open waters cause environmental pollution, which, as a chain reaction, trigger changes in the natural habitats of flora and fauna [1\u20133]. Moreover, accumulation of CO2, N2O, CH4 and SO2 gases in the atmosphere cause health problems for millions of people and accelerate climate change [4].\n\nTo sustain a global population of 7.7 billion people and manage their environmental footprint sustainably, the industry should transition towards a circular economy by using renewable resources and implementation of sustainable technologies. One such technology is the gas biodesulfurization process developed by our group in the Department of Environmental Technology at Wageningen University in cooperation with Delft University of Technology, University of Amsterdam, and industrial partners: Paqell B.V, Paques B.V and Shell. This technology emerged in the early 1990s when physicochemical desulfurization processes were dominating the market. Our biodesulfurization technology distinguishes itself because of its reduced operational and capital expenditures, and smaller environmental footprint. Since then, gas biodesulfurization has been intensively studied in order to facilitate higher sulfur recovery rates (>90 mol%) and stable process operations while treating a variety of gas feed streams. A high selectivity for sulfur is preferred because this will regenerate hydroxide ions, which are consumed when H2S is removed from gas streams. In addition, the consumption of air, energy, and caustic at sulfur-producing conditions are lower relative to the formation of sulfuric acid [5]. Furthermore, the recovered sulfur slurry can be used as fertilizer and as fungicide [6]. To maintain a stable sulfur selectivity, the biodesulfurization process operation should remain stable as well, especially when gas feed composition and sulfide concentration fluctuate.\n\nThe composition of the feed gas depends on the industry that generates the sour gas. For example, biogas formed from the anaerobic digestion of wastewater in paper mill facilities has a relatively low amount of H2S (0.7 vol.%), whereas sour gas streams in the oil and gas industry are composed of up to 95 vol.% of H2S, a fraction of CO2, hydrocarbons, and thiols. The H2S concentration can vary greatly, not only between industries but also during the operation of a single installation. The daily H2S loading rate between Thiopaq installations may range from 10 kg day-1 up to 150 ton day-1. Therefore, the aim of this research was to achieve more sulfur formation and stable process operation in the presence of thiols; see Chapter 1 for main research questions. In the following section we will discuss which measures can be applied to increase process stability and sulfur formation under variations in gas feeds.\n\n7.2. Achieving optimal process performance\n\nIn the studies described in this thesis, we made use of a variety of scientific disciplines ranging from molecular biology, toxicology, process technology to process modeling and process control in order to better understand the four key elements required for achieving optimal process requirements. These elements are described below:\n\n1. Development of a methodology for selecting SOB biomass to start up a full-scale gas biodesulfurization installation. SOB have two enzymatic routes for sulfide oxidation that have been known so far. The majority of SOB were found to use flavocytochrome c oxidase to transfer electrons to cytochrome c. This route is called FCC that converts HS- to S8. The second route that is used by some SOB species, i.e. Alkalilimnicola, is sulfide-quinone reductase (SQR) that uses quinones as electron carriers. It was found that SQR pathway is more energetically favorable and less sensitive to toxic compounds such as thiols [7]. Furthermore, Klok and colleagues found that SQR route prevails when sulfide oxidation takes place under oxygen limiting conditions and stimulates elemental sulfur formation [8]. Therefore, finding ratio between expression rates of sulfide oxidation enzyme routes FCC and SQR will enable to understand capacity of the chosen SOB form sulfur and sulfate. This FCC\/SQR ratio is derived from the respiration rate data and introduced as an \u03b1 value. In our study, we found that at \u03b1 < 0.8, sulfur formation prevailed in the process, whereas product formation shifted towards sulfate formation at \u03b1 > 0.8 (Chapter 2). This dependency between \u03b1, sulfate and sulfur formation can be used as an assessment tool for a process engineer\/installation operation to assess product formation rates and develop debottlenecking strategies for full-scale installations. In addition, it can be used to evaluate a chosen inoculum on its potential for sulfur and sulfate formation before starting the process.\n\n2. Evaluation of the dual biodesulfurization line-up in the presence and absence of thiols by investigating underlying biochemical reactions of H2S oxidation and SOB community dynamics. In our study, we achieved a sulfur selectivity of ~92 mol% in both the double-bioreactor setup and the traditional line-up during short-term experiments of about 2 weeks of operation (Chapters 2 and 4). However, we noticed that the addition of a second, anaerobic, bioreactor to the line-up resulted in the increase of the process operability as both formation rates and ORP values were more stable in that setup. It appeared that the added anaerobic bioreactor in the double-bioreactor line-up was damping fluctuations of the sulfide-rich solution flow into the aerobic bioreactor. The ORP sensor readings showed that the values measured in anaerobic bioreactor were consistently more negative than in the aerobic bioreactor (~-430 mV vs. ~-390 mV), meaning that no oxygen was present in the anaerobic bioreactor and that conditions in the anaerobic bioreactor were more reduced. This difference in the ORP values and oxygen availability affects the state of bacterial cells, more importantly state of the enzymatic routes for biological sulfide oxidation [9]. Moreover, the addition of an anaerobic bioreactor increased the biomass retention time and the interaction of SOB with organic and inorganic sulfur compounds. These improvements prompted us to study the effect of thiols on the process operation of the double bioreactor setup (Chapter 5). In this work, we identified that almost all sulfide supplied to the system in the feed gas was in the form of polysulfide. This indicates that, in our experimental setup, the dominating electron donor for SOB was polysulfide. Moreover, we found that when the double bioreactor was added to the process line-up, we were able to increase the selectivity for sulfur formation up to 90 mol% when methanethiol (MT) was supplied at a loading rate of 2 mM S day-1, whereas in the traditional line-up, the maximum for sulfur formation was 75% [10].\n\nIn addition to the above, we observed that the anaerobic bioreactor provided selective pressure on the SOB microbial community. To assess SOB dynamics, we initially performed 16S rRNA gene amplicon analyses on samples collected at equal time intervals during experiments with and without MT addition (Chapters 4 and 5). The output of amplicon sequencing provides the relative abundance of microbial species present in the community at the moment of sampling. These abundances are relative as they depend on an abundance of other species within the SOB community. Therefore, in addition, a quantitative assay was performed to answer more profound questions on species interactions, dependency of the process performance and growth dynamics. We cloned our sludge and identified the three most dominant SOB species: Thioalkalivibrio sulfidiphilus, Alkalilimnicola ehrlichii, and Thioalkalibacter halophilus. For these species, we developed species-specific primers and qPCR assay (Chapter 3). The qPCR results show that proliferation of Thioalkalibacter halophilus and decrease of Thioalkalivibrio sulfidiphilus was strongly correlated to the presence of MT and\/or its oxidation product dimethyl disulfide (DMDS). IC50 values of Thioalkalibacter halophilus confirmed its high tolerance to dimethyl disulfide (IC50 value at 2.37 mM of DMDS) (Chapter 4).\n\n3. Selective inhibition of sulfate formation by DMDS addition. The idea for this experiment originated from the findings of Roman et al. (2016b), describing the inhibition modes of organic sulfur compounds. In this study, we found that DMDS inhibits the oxidation of internal poly sulfur compounds into sulfate. The first experimental run showed a strong reduction of sulfate formation. The quantified sulfate selectivity of about ~ 1 mol% was reduced with ~7 mol% compared with sulfate formation in the absence of DMDS (Chapter 4). However, no differences were found for sulfur formation in the absence or presence of DMDS, i.e. both runs showed selectivity for sulfur formation of ~90 mol%. The remaining sulfide was chemically converted to thiosulfate at ~9 mol%. Chemical oxidation rates of dissolved sulfide increased when biological sulfide oxidation rates were suppressed. To prove that thiosulfate formation can be reduced by increasing biological activity, we performed a test with a doubled biomass concentration. We achieved 0 mol% thiosulfate formation with ~96 mol% sulfur formation. In Chapter 4, we also presented a new method developed in-house for the quantification of organic sulfur compounds in liquid samples. With this method, we could identify diorgano polysulfanes which formed in the process medium. The identification of formed diorgano polysulfanes resulted in the proposition of the novel reaction between sulfide and dimethyl disulfide, which leads to the formation of longer-chain diorgano polysulfanes, i.e., dimethyl trisulfide and dimethyl tetrasulfide.\n\n4. Development of an alternative oxygen supply control strategy to enable stable process operation with high sulfur selectivity (>95 mol%) in the presence of thiols. Most of the industrial processes use automated control to steer product formation and operational parameters such as pH, dissolved oxygen, alkalinity, and ORP. In the current gas biodesulfurization process, an integrated PID feedback-based controller is used, which regulates the air supply based on input signals from the ORP sensor. To increase the sulfur formation rate and to overcome fluctuations of H2S and thiols concentration in the feed gas, we developed an alternative, feedforward control strategy for oxygen supply (Chapter 6). This strategy is based on a fixed O2\/H2S supply ratio to the aerated bioreactor. In the past, this ratio was used by our research group to assess the system\u2019s performance. In our study, we used this parameter as a control variable. Based on the available set of previous studies, we knew that the highest selectivity for sulfur formation was achieved at an O2\/H2S supply ratio of ~0.60 mol mol-1 [5,12,13]. Therefore, we selected a control value for the feedforward control of ~0.60 mol mol-1. The results of our experiments with feedforward control show stable process performance and high sulfur selectivity of ~96 mol% at the randomly varying H2S supply. Preliminary results also show that the high sulfur formation rate of ~95 mol% is not affected by the presence of ethanethiol (1.16 mM S day-1).\n\nBy the completion of all experimental runs, we found out that process operation in terms of sulfur selectivity was higher than 90 mol%, when lab-scale set-up was inoculated with SOB biomass from the full-scale facilities that operate well. Chapter 2 shows differences in process operation with four SOB biomasses taken from four different full-scale installations. Sulfur selectivity was higher when the SOB biomass was taken from a full-scale installation that was operating well. By \u201coperating well\u201d, we mean stable process operation for a prolonged period of time with a sulfur selectivity of ~90 mol%. Moreover, SOB composition was also found to determine process performance. Therefore, to be able to optimize process performance, the SOB community should be able to withstand feed gas composition. For instance, in the experiments with addition of DMDS (Chapter 4) and MT (Chapter 5) the inoculum consisted of a mix of SOB from various installations: 40% was from the Oilfield - 1 installation, which treats associated gas with low concentrations of thiols 50-200 ppm and 1-5% of H2S. Another 40% was SOB biomass from a pilot plant, which treats the synthetically prepared gas that represents amine acid gas with 4.45% of H2S where a high abundance of Alkalilimnicola ehrlichii was found [14]. 10% was from Oilfield \u2013 2 installations located in South-East Asia where the feed gas contains next to H2S also a high concentration of thiols. The final 10% was from an installation treating landfill gas, with a SOB biomass showed high sulfur selectivity in the presence of sulfide (Chapter 2). By preparing a mixture of the inocula, we increased the chances of developing a SOB community that would be able to oxidize sulfide in the presence of MT and DMDS. After enrichment, we identified two important species that are vital to enable stable process performance in the presence of MT: Thioalkalibacter halophilus and Alkalilimnicola ehrlichii. By contrast, in the absence of MT, Thioalkalivibrio sulfidiphilus species are essential to enable stable process operation.\n\n7.3. Identified knowledge gaps\n\nThroughout 30 years of research by our group, numerous studies have been performed to understand and optimize the biological desulfurization process. Many analytical tools and microbiological identification methods have been developed to analyze biological systems that are now commercially available. Therefore, new questions can be addressed and studied in more depth, such as the ecophysiology of the species within the SOB community and which enzyme system is used by the SOB for sulfide oxidation in the presence and absence of VOSCs. Furthermore, it might be possible to find SOB species that are not only able to withstand thiols but also perform conversion of thiols.\n\nAnother issue that requires more investigations is the robustness and up-scaling of the developed feedforward control. In the described research, we observed significant differences between feedback and feedforward control with fluctuating H2S load. Feedforward control could be beneficial in gas biodesulfurization systems to achieve better operational stability and potentially higher sulfur selectivity in the presence of thiols. Therefore, further studies should aim at investigating the robustness of feedforward control at the elevated thiol loading rates. In addition, a feedforward control strategy should be implemented in a pilot installation and perform fully automated control by quantification of H2S in the gas feed. As a first step, H2S sensors need to be screened and tested for high robustness and reliability.\n\n7.3.1 Ecophysiology of the species within SOB community\n\nIn bioprocesses, the emphasis is often centered on product formation and on maximizing yields or recovery factors of the compounds of interest. However, major changes occur within, especially, mixed bacterial communities. Furthermore, it is essential to link microbial activity and engineering performance, as well as to gain knowledge on the physiology of the species makeup of the SOB community as it enables optimal use of the community in bioprocesses [15]. Nowadays, bioprocess engineers usually perform 16S rRNA amplicon sequencing to determine community structure and to associate process performance with the generally most abundant species. However, the dominant organisms may not necessarily play the critical role in the community structure as well as in product formation [16]. Therefore, only performing 16S rRNA analysis is not enough to answer more profound questions such as why certain products were formed, and why one bacterial species proliferates at certain conditions.\n\nIn the era of genomics, various assays have been already developed (Table 1) that can help us understand the dynamics and structure of SOB communities under certain process conditions [17]. Moreover, these -omics assays can provide crucial information on the sulfur metabolism of each species in the SOB community, as we may not even know what these species can do more of. Currently, the S-metabolism of only a few SOB species is known, and more importantly what their capacities are and what those species actually do in the gas biodesulfurization process. Therefore, the S-metabolism of other SOB species in the gas biodesulfurization process needs to be unraveled as well, to gain a better understanding of which bacteria are responsible for which activity. This will enable us to tailor the inoculum to fit the gas feed composition and facilitate higher process efficiencies. For instance, in our research, we found two bacterial species highly abundant in the presence of MT and DMDS: Thioalkalibacter halophilus and Alkalilimnicola ehrlichii. However, little is known about their role in gas biodesulfurization, their S-metabolism, oxidation capacity and why they are able to thrive in the presence of MT. What makes the number of other species decrease drastically in the presence of MT? That is why more insight is required into the genomes of SOB species and their -omics.\n\nTable 1 Various -omics technologies to assess functionality of the bacterial community.\n- Metagenomics (DNA): assesses the density of microbial communities and their genetic and functional diversity [19].\n- Metatranscriptomics (RNA): concentrates on expressed genes in the entire microbial community and provides a view of the active functions of the community of microorganisms [20].\n- Metaproteomics (Protein): assesses the \u201cexpressed\u201d metabolism and physiology of microbial community members, helping to understand the functioning of the ecosystem [21].\n- Metabolomics (Metabolites): provides information on the metabolites (composition), which helps to understand the functional dynamics influencing community and host interactions.\n\n7.3.2 Development of molecular tools to monitor expression of FCC and SQR enzymes\n\nApart from using -omics assays, which are time-consuming and comparatively costly, real-time PCR (RT-PCR) allows fast analyses of environmental samples on singular or multiple gene expression [18]. The genes of interest are the ones that encode the flavocytochrome c oxidase (FCC) and sulfide-quinone oxidoreductase (SQR) pathways that are used by SOB for sulfide oxidation (Fig. 1). It is of high importance to understand which enzymatic route SOB species are using in the presence of thiols. This may also provide information on the mechanism of detoxification.\n\nFig. 1. Schematic representation of the chemolithotrophic electron transport chain from sulfide during bacterial respiration (adapted from [22]) flavocytochrome c and sulfide-quinone reductase. Sulfide-quinone reductase (SQR).\n\n7.3.3 Robustness of the feedforward control at high thiol loadings\n\nThe presented feedforward control shows promising results for the gas biodesulfurization process control resulting in an increase in sulfur formation of to 96 mol%. However, before implementing feedforward control in full-scale installations, it needs to be implemented and tested in a larger pilot installation. In addition, the robustness of the feedforward control needs to be tested with addition of thiols, BTEX and other compounds that are present in the gas feed in full-scale installations.\n\nIn addition, it might be possible to combine feedforward control into model predictive control (MPC). The idea is to pair sulfide readings from the gas feed, with a computer, and have a software perform in-line calculations of the required oxygen supply based on the set O2\/H2S ratio limits. The oxygen\/air valve could then be controlled this way, and oxygen supplied to the aerobic bioreactor as needed. In addition to setting O2\/H2S ratio limits, it might be possible to integrate MPC with control objective prioritization and symptom-aided diagnosis [23,24].\n\nFig. 2. Schematic representation of the basic structure of model predictive control.","auteur":"Karine Kiragosyan","auteur_slug":"karine-kiragosyan","publicatiedatum":"8 april 2020","taal":"EN","url_flipbook":"https:\/\/ebook.proefschriftmaken.nl\/ebook\/karinekiragosyan?iframe=true","url_download_pdf":"https:\/\/ebook.proefschriftmaken.nl\/download\/97900a15-d922-4b60-83d6-f6019b5dbc7a\/optimized","url_epub":"","ordernummer":"FTP-202604031254","isbn":"978-94-6395-258-3","doi_nummer":"","naam_universiteit":"Wageningen University","afbeeldingen":14030,"naam_student:":"","binnenwerk":"","universiteit":"Wageningen University","cover":"","afwerking":"","cover_afwerking":"","design":""},"_links":{"self":[{"href":"https:\/\/www.proefschriftmaken.nl\/en\/wp-json\/wp\/v2\/us_portfolio\/8251","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.proefschriftmaken.nl\/en\/wp-json\/wp\/v2\/us_portfolio"}],"about":[{"href":"https:\/\/www.proefschriftmaken.nl\/en\/wp-json\/wp\/v2\/types\/us_portfolio"}],"author":[{"embeddable":true,"href":"https:\/\/www.proefschriftmaken.nl\/en\/wp-json\/wp\/v2\/users\/8"}],"replies":[{"embeddable":true,"href":"https:\/\/www.proefschriftmaken.nl\/en\/wp-json\/wp\/v2\/comments?post=8251"}],"version-history":[{"count":1,"href":"https:\/\/www.proefschriftmaken.nl\/en\/wp-json\/wp\/v2\/us_portfolio\/8251\/revisions"}],"predecessor-version":[{"id":8254,"href":"https:\/\/www.proefschriftmaken.nl\/en\/wp-json\/wp\/v2\/us_portfolio\/8251\/revisions\/8254"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.proefschriftmaken.nl\/en\/wp-json\/wp\/v2\/media\/14030"}],"wp:attachment":[{"href":"https:\/\/www.proefschriftmaken.nl\/en\/wp-json\/wp\/v2\/media?parent=8251"}],"wp:term":[{"taxonomy":"us_portfolio_category","embeddable":true,"href":"https:\/\/www.proefschriftmaken.nl\/en\/wp-json\/wp\/v2\/us_portfolio_category?post=8251"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}