Steam reforming of methanol for ultra-pure H2 production in a membrane reactor: Techno-economic analysis

Process simulation and design as well as economic analysis were carried out to evaluate technical and economic feasibility of steam reforming of methanol in a membrane reactor (MR) for ultra-pure H₂ production. Using a commercial process simulator, Aspen HYSYS®, comparative studies were conducted to investigate the effect of operating conditions including the H₂ permeance (1 × 10⁻⁵ - 6 × 10⁻⁵ mol m⁻² s⁻¹ Pa⁻¹), a H₂O sweep gas flow rate (1–20 kmol h⁻¹), and a reaction temperature (448–493 K) in a conventional packed-bed reactor (PBR) and the MR using a previously reported reaction kinetics. Improved performances such as methanol conversions and H₂ yields were observed in the MR compared to the PBR and several design guidelines for the MR were obtained to develop H₂ separation membranes with optimal H₂ permeance and to select a suitable H₂O sweep gas flow rate. In addition, economic analysis based on itemized cost estimations was conducted for a small-sized H₂ fueling station by calculating a unit H₂ production cost for both the PBR and the MR reflecting a current economic status in Korea. As a result, a cost saving of about 23% was obtained in the MR (7.24 $ kgH₂⁻¹) compared to the PBR (9.37 $ kgH₂⁻¹) confirming the benefit of employing the MR for ultra-pure H₂ production.     

Cost‐competitive methane steam reforming in a membrane reactor for H2 production: Technical and economic evaluation with a window of a H2 selectivity

Techno‐economic viability studies of employing a membrane reactor (MR) equipped with H₂ separation membranes for methane steam reforming (MSR) were carried out for H₂ production in Korea using HYSYS®, a well‐known chemical process simulator, including economic analysis based on itemized cost estimation and sensitivity analysis (SA). With the reaction kinetics for MSR reported by Xu and Froment, the effect of a wide range of H₂ selectivity (10‐10,000) on the performance in an MR was investigated in this study. Because of the equilibrium shift owing to the Le Chatelier's principle, great performance of enhancement of methane conversion ( ) and H₂ yield and reaction temperature reduction was observed in an MR compared with a packed‐bed reactor (PBR). A window of a H₂ selectivity from 100 to 300 is proposed as a new criterion for better MR performance of MSR depending on potential applications from in‐depth analysis of and H₂ yield enhancements, a H₂ purity, and temperature reduction. In addition, economic analysis to evaluate the feasibility of an MR technology for MSR was carried out focusing on a levelized cost of H₂ based on itemized cost estimation of capital and operating costs as well as SA. Techno‐economic assessment showed 36.7% cost reduction in an MR compared with a PBR and revealed that this MR technology can be possibly opted for a cost‐competitive H₂ production process for MSR.     

Techno-economic analysis (TEA) for CO2 reforming of methane in a membrane reactor for simultaneous CO2 utilization and ultra-pure H2 production

Techno-economic analysis (TEA) for CO₂ reforming of methane in a membrane reactor (MR) was conducted by using process simulation and economic analysis. Parametric studies for key operating conditions like a H₂ permeance, a H₂O sweep gas flow rate, operating temperature, and a CO₂/CH₄ ratio were carried out for a conventional packed-bed reactor (PBR) and a MR using Aspen HYSYS^®, a commercial process simulator program and some critical design guidelines for a MR in terms of a H₂O sweep gas flow rate and a CO₂/CH₄ ratio were obtained. Further economic analysis based on process simulation results showed about 42% reduction in a unit H₂ production cost in a MR (6.48 $ kgH₂^?1) than a PBR (11.18 $ kgH₂^?1) mostly due to the elimination of a pressure swing adsorption (PSA) system in a MR. In addition, sensitivity analysis (SA) revealed that reactant price and labor were the most influential economic factors to determine a unit H₂ production cost for both a PBR and a MR. Lastly, profitability analysis (PA) from cumulative cash flow diagram (CCFD) in Korea provided positive net present value (NPV) of $443,760～$240,980, discounted payback period (DPBP) of 3.03–3.18 y, and present value ratio (PVR) of 7.51–4.97 for discount rates from 2 to 10% showing economic feasibility of the use of a MR as simultaneous CO₂ utilization and ultra-pure H₂ production.     

Improved hydrogen production of the downstream bioreactor by coupling single chamber microbial fuel cells between series-connected photosynthetic biohydrogen reactors

To obtain additional hydrogen recovery from the downstream photosynthetic biohydrogen reactor (PBR), a system (PBR1–MFCs–PBR2) that combined PBRs with three single chamber microbial fuel cells (MFCs) was proposed in this study. The results revealed that the PBR2 in PBR1–MFCs–PBR2 showed a hydrogen production rate of 0.44 ± 0.22 mmol L h⁻¹, which was 15 and 4 times higher than those obtained by direct connecting the two PBRs (PBR1–PBR2) and pH regulated system (PBR1–pH regulation–PBR2), respectively. In addition, the PBR1–MFCs–PBR2 exhibited the highest glucose utilization (η_g) of 97.6 ± 2.1 %, while lower η_g values of 75.6 ± 2.2% and 70.1 ± 1.2% was obtained for PBR1–PBR2 and PBR1–pH regulation–PBR2, respectively. These improvements were due to the removal of inhibitory byproduct and H⁺ from the PBR1 effluent by the MFCs.     

H2 production by low pressure methanol steam reforming in a dense Pd–Ag membrane reactor in co-current flow configuration: Experimental and modeling analysis

The aim of this work is to generate a pure or CO_x-free hydrogen stream by using a dense Pd-based packed bed membrane reactor (PBMR) during methanol steam reforming (MSR) reaction and developing a valid model that can provide a tool for deeper analyses of the reaction parameters in the PBMR. Therefore, in this study, a dense Pd–Ag membrane reactor (MR) is used to carry out MSR at different gas hourly space velocity (GHSV), feed molar ratio and sweep gas factor (SF) and for low reaction pressures (1.5–2.5 bar). For a better analysis, a traditional packed bed reactor (PBR) is operated at the same PBMR conditions. In the PBMR setup, a dense Pd–Ag membrane with a thickness of 50 μm is used and also a commercial Cu/ZnO/Al₂O₃ catalyst is packed in both kinds of reactors. Methanol conversion equal to 100% is experimentally achieved in the PBMR at 280 °C, H₂O/CH₃OH = 3/1 and 2.5 bar, while at the same conditions the PBR reaches 91% methanol conversion. Moreover, 46% CO_x-free hydrogen on total hydrogen produced is collected by using sweep gas in the PBMR permeate side. Furthermore, a 1-dimensional and isothermal model is developed for theoretically analyzing MSR performance in both PBMR and PBR, validated by the combined experimental campaign.     

Production of hydrogen from methanol steam reforming using CuPd/ZrO2 catalysts – Influence of the catalytic surface on methanol conversion and CO selectivity

Electricity generation for mobile applications by proton exchange membrane fuel cells (PEMFCs) is typically hindered by the low volumetric energy density of hydrogen. Nevertheless, nearly pure hydrogen can be generated in-situ from methanol steam reforming (MSR), with Cu-based catalysts being the most common MSR catalysts. Cu-based catalysts display high catalytic performance, even at low temperatures (ca. 250 °C), but are easily deactivated. On the other hand, Pd-based catalysts are very stable but show poor MSR selectivity, producing high concentrations of CO as by-product. This work studies bimetallic catalysts where Cu was added as a promoter to increase MSR selectivity of Pd. Specifically, the surface composition was tuned by different sequences of Cu and Pd impregnation on a monoclinic ZrO₂ support. Both methanol conversion and MSR selectivity were higher for the catalyst with a CuPd-rich surface compared to the catalyst with a Pd-rich surface. Characterization analysis indicate that the higher MSR selectivity results from a strong interaction between the two metals when Pd is impregnated first (likely an alloy). This sequence also resulted in better metallic dispersion on the support, leading to higher methanol conversion. A H₂ production rate of 86.3 mmol h^?1 g^?1 was achieved at low temperature (220 °C) for the best performing catalyst.     

Evaluation of continuous biohydrogen production from enzymatically treated cornstalk hydrolysate

Enzymatically treated cornstalk hydrolysate was tested as substrate for H₂ production by Thermoanaerobacterium thermosaccharolyticum W16 in a continuous stirred tank reactor. The performance of strain W16 to ferment the main components of hydrolysate, mixture of glucose and xylose, in continuous culture was conducted at first, and then T. thermosaccharolyticum W16 was evaluated to ferment fully enzymatically hydrolysed cornstalk to produce H₂ in continuous operation mode. At the dilution rate of 0.020 h⁻¹, the H₂ yield and production rate reached a maximum of 1.9 mol H₂ mol⁻¹ sugars and 8.4 mmol H₂ L⁻¹ h⁻¹, respectively, accompanied with the maximum glucose and xylose utilizations of 86.3% and 77.6%. Continuous H₂ production from enzymatically treated cornstalk hydrolysate in this research provides a new direction for economic, efficient, and harmless H₂ production.     

Comparative kinetic modeling of growth and molecular hydrogen overproduction by engineered strains of Thermotoga maritima

Thermotoga maritima is an anaerobic hyperthermophilic bacterium known for its high amounts of hydrogen (H₂) production. In the current study, the kinetic modeling was applied on the engineered strains of T. maritima that surpassed the natural H₂ production limit. The study generated a kinetic model explaining H₂ overproduction and predicted a continuous fermentation system. A Leudking-Piret equation-based model predicted that H₂ production by Tma200 (0.217 mol-H₂ g⁻¹-biomass) and Tma100 (0.147 mol-H₂ g⁻¹-biomass) were higher than wild type (0.096 mol-H₂ g⁻¹ -biomass) with reduced rates of maltose utilization. Sensitivity analysis confirmed satisfactory fitting of the experimental data. The slow growth rates of Tma200 (0.550 h⁻¹) and Tma100 (0.495 h⁻¹) are compared with the wild type (0.663 h⁻¹). A higher maintenance energy along with growth and non-growth H₂ coefficients corroborate the higher H₂ productivity of the engineered strains. The modeled data established a continuous fermentation system for the sustainable H₂ production.     

Sustained photobiological hydrogen production by Chlorella vulgaris without nutrient starvation

This article describes the ability of the Chlorella vulgaris BEIJ strain G-120 to produce hydrogen (H₂) via both direct and indirect pathways without the use of nutrient starvation. Photobiological H₂ production reached a maximum rate of 12 mL H₂ L^?1 h^?1, corresponding to a light conversion efficiency (light to H₂) of 7.7% (average 3.2%, over the 8-day period) of PAR, (photosynthetically active irradiance). Cells presented a maximum in vivo hydrogenase activity of 25.5 ± 0.2 nmoles H₂ μgChl^?1 h^?1 and the calculated in vitro hydrogenase activity was 830 ± 61 nmoles H₂ μgChl^?1 h^?1. The strain is able to grow either heterotrophically or photo autotrophically. The total output of 896 mL of H₂ was attained for illuminated culture and 405 mL for dark cultures. The average H₂ production rate was 4.98 mL L^?1 h^?1 for the illuminated culture and 2.08 mL L^?1 h^?1 for the one maintained in the dark.     

Enzymatic routes to hydrogen and organic acids production from banana waste fermentation by autochthonous bacteria: Optimization of pH and temperature

The Central Composite Rotational Design (CCRD) was employed to find the optimum pH (5.09–7.91) and temperature (27.1–46,9 °C) for hydrogen production in banana waste (BW) fermentation by autochthonous microbial biomass. The P and Rm ranged between 6.06 and 62.43 mL H₂ and 1.13–12.56 mL H₂.h^?1, respectively. The temperature 37 °C and pH 7.0 were the optimum conditions for P (70.19 mL H₂) and Rm (12.43 mL H₂.h^?1) as predicted by the mathematical model. Fructose and glucose are the primary alternative carbon sources in banana waste-fed batch reactors. The high concentration of lactic acid and H₂ production was associated to Lactobacillus (52–81%) and Clostridium (14–35%). However, the most important finding was about butyric acid (HBu). This acid is the better indicator of hydrogen production than acetic acid (HAc). The pH effected carbohydrates fermentation and organic acids production. The genes encoding the enzymes related to galactose, sucrose, fructose, arabinose and xylose metabolism were predominant.     

Lignocellulosic ethanol production employing immobilized Saccharomyces cerevisiae in packed bed reactor

Most of ethanol production processes are limited by lower ethanol production rate and recyclability problem of ethanologenic organism. In the present study, immobilized co-fermenting Saccharomyces cerevisiae GSE1618 was employed for ethanol fermentation using rice straw enzymatic hydrolysate in a packed bed reactor (PBR). The immobilization of S. cerevisiae was performed by entrapment in Ca-alginate for optimization of ethanol production by varying alginic acid concentration, bead size, glucose concentration, temperature and hardening time. Remarkably, extra hardened beads (EHB) immobilized with S. cerevisiae could be used up to repeated 40 fermentation batches. In continuous PBR, maximum 81.82 g L⁻¹ ethanol was obtained with 29.95 g L⁻¹ h⁻¹ productivity with initial glucose concentration of 180 g L⁻¹ in feed at dilution rate of 0.37 h⁻¹. However, maximum ethanol concentration of 40.33 g L⁻¹ (99% yield) with 24.61 g L⁻¹ h⁻¹ productivity was attained at 0.61 h⁻¹ dilution rate in fermentation of un-detoxified rice straw enzymatic hydrolysate (REH). At commercial scale, EHB has great potential for continuous ethanol production with high productivity using lignocellulosic hydrolysate in PBR.     

Hydrogen-producing purple non-sulfur bacteria isolated from the trophic lake Averno (Naples,Italy)

Seventeen purple non-sulfur bacterial strains, isolated from the trophic lake Averno, Naples, Italy, were phylogenetically classified and their H₂-producing performances were tested utilizing various synthetic substrates and the fermentation broth derived from the spontaneous fermentation of vegetable residues. All the strains showed the capability to produce hydrogen on at least one of the four carbon substrates tested (malic, lactic, acetic and succinic acid). On lactate, Rhodopseudomonas palustris strain AV33 showed the best maximum production rate (50.7 ± 2.6 mL (H₂) L⁻¹ h⁻¹), with a mean rate, calculated on the whole period of production, of 17.9 mL ± 0.7 (H₂) L⁻¹ h⁻¹. In the presence of acetate, AV33 produced only few mL of H₂, but intracellularly accumulated poly-β-hydroxybutyrate up to a concentration of 21.4 ± 3.4% (w/w) of cell dry weight. Rp. palustris AV33 also produced H₂ on the fermentation broth supplemented with Fe, with a maximum production rate of 16.4 ± 2.3 mL (H₂) L⁻¹ h⁻¹ and a conversion yield of 44.2%.     

Mechanistic insights into methanol steam reforming over a ZnZr oxide catalyst with improved activity

Methanol steam reforming (MSR) holds great potential for mobile hydrogen production, but it still requires an active and stable catalyst. In this work, we report a high-performance ZnZr-0.5 composite oxide catalyst for this reaction, with a hydrogen production rate of 2.80 mol·g_cat^?1·h^?1 and CO₂ selectivity of 99.6% at a methanol space velocity of 22,762 mL·g_cat^?1·h^?1. It also exhibits superior long-term durability in the TOS test for more than 100 h. Such good activity results from a synergistic effect of ZnO–ZrO₂ dual sites. ZrO₂ is capable of stabilizing and storing more CH₃O1 and HCOO1 intermediates while ZnO is in charge of the dehydrogenation of these key intermediates. In situ diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) and chemisorption results reveal that the MSR reaction experiences successively the hydrolysis of methyl formate and dehydrogenation of formate. More importantly, it is found that H₂O significantly promotes the dehydrogenation of HCOO1 intermediate by directly participating in this reaction from pulse chemisorption experiments.     

Improvement of hydrogen production by newly isolated Thermoanaerobacterium thermosaccharolyticum IIT BT-ST1

Efficient H₂ producing bacterial strain Thermoanaerobacterium thermosaccharolyticum IIT BT-ST1 was isolated from the anaerobic digester. Taguchi design of experiment was applied to evaluate the influence of the temperature, pH, glucose, FeSO₄ and yeast extract on H₂ production with three levels of orthogonal array in the experimental design. Temperature showed most significant influence on the H₂ production process. Investigation of mutual interaction between the process parameters was studied employing Box–Behnken design. Experimentally optimized process parameters (60 °C, pH 6.5, 20 mM FeSO₄, 4 g L⁻¹ yeast extract and 12 g L⁻¹ glucose) gave the maximum H₂ production of 3930 mL L⁻¹ in 24 h, which have close resemblance with the theoretical values. Continuous H₂ production using packed bed reactor was studied. Maximum H₂ production rate of 1691 mL L⁻¹ h⁻¹ at a dilution rate of 0.6 h⁻¹ was observed which is about 10 times higher than the batch process.     

Fed-batch operation for bio-H2 production by Rhodopseudomonas palustris (strain 42OL)

Interest in renewable and clean energies such as hydrogen has increased because of the high level of polluting emissions, increasing costs associated with petroleum and the escalating problems of global climate change. In the presence of a light source, a microbial photosynthetic process provides a system for the conversion of some organic compounds into biomass and hydrogen. Using Rhodopseudomonas palustris as a cell-factory, hydrogen photo-evolution was investigated in a photobioreactor (PBR) irradiated either from one or two opposite sides. Irradiating the photobioreactor from only one side, in the presence of malic acid, a reactor hydrogen production of 2.786 l(H₂) PBR⁻¹ was achieved. When the PBR was irradiated from two opposite sides, hydrogen photo-evolution increased to 3.162 l(H₂) PBR⁻¹. Experiments were carried out using inoculum from either the retardation or the exponential growth phases. Using the latter, the highest hydrogen photo-evolution rate based on the bacteriochlorophyll (Bchl) concentration was achieved (3295 μl(H₂) mg (Bchl⁻¹ h⁻¹). The hydrogen to biomass ratio (r_g) was 1.91 l g⁻¹ in the medium containing malic acid and 1.07 l g⁻¹ in that containing acetic acid. It was found that the hydrogen production rate was higher with malic than with acetic acid. Although photobiological hydrogen production cannot furnish alone the greater and greater world requirements of clean renewable energy, it is desirable that photobiological hydrogen technology will grow, in the near future, because photobioreactors for bio-hydrogen production can be positioned in fringe areas without competition with agricultural lands.     

Economic evaluation with uncertainty analysis using a Monte-Carlo simulation method for hydrogen production from high pressure PEM water electrolysis in Korea

Economic analysis with uncertainty analysis based on a Monte-Carlo simulation method was performed for hydrogen production from high pressure PEM water electrolysis targeting a hydrogen production capacity of 30 Nm³ h^?1 in Korea. With key economic parameters obtained from sensitivity analysis (SA), a cumulative probability curve was constructed for a unit H₂ production cost fully reflecting unpredictable price fluctuations in H₂ production equipment, construction, electricity, and labor from ±10% to ±50%. In addition, economic analysis for a net present value (NPV) with uncertainty analysis for revenue (REV), fixed capital investment (FCI), and cost of manufacturing (COM) provided cumulative probability curves with different discount rates and more reliable NPVs (-$ 69,000 ～$1,308,000) for high pressure PEM water electrolysis under development in Korea. This economic analysis based on uncertainty can serve as important economic indicators suitable for premature technology like high pressure PEM water electrolysis currently being in progress in Korea.     

Hydrogen production from acid hydrolyzed molasses by the hydrogen overproducing Escherichia coli strain HD701 and subsequent use of the waste bacterial biomass for biosorption of Cd(II) and Zn(II)

This study was devoted to investigate production of hydrogen gas from acid hydrolyzed molasses by Escherichia coli HD701 and to explore the possible use of the waste bacterial biomass in biosorption technology. In variable substrate concentration experiments (1, 2.5, 5, 10 and 15 g L⁻¹), the highest cumulative hydrogen gas (570 ml H₂ L⁻¹) and formation rate (19 ml H₂ h⁻¹ L⁻¹) were obtained from 10 g L⁻¹ reducing sugars. However, the highest yield (132 ml H₂ g⁻¹ reducing sugars) was obtained at a moderate hydrogen formation rate (11 ml H₂ h⁻¹ L⁻¹) from 2.5 g L⁻¹ reducing sugars. Subsequent to H₂ production, the waste E. coli biomass was collected and its biosorption efficiency for Cd²⁺ and Zn²⁺ was investigated. The biosorption kinetics of both heavy metals fitted well with the pseudo second-order kinetic model. Based on the Langmuir biosorption isotherm, the maximum biosorption capacities (q_max) of E. coli waste biomass for Cd²⁺ and Zn²⁺ were 162.1 and 137.9 (mg/g), respectively. These q_max values are higher than those of many other previously studied biosorbents and were around three times more than that of aerobically grown E. coli. The FTIR spectra showed an appearance of strong peaks for the amine groups and an increase in the intensity of many other functional groups in the waste biomass of E. coli after hydrogen production in comparison to that of aerobically grown E. coli which explain the higher biosorption capacity for Cd²⁺ or Zn²⁺ by the waste biomass of E. coli after hydrogen production. These results indicate that E. coli waste biomass after hydrogen production can be efficiently used in biosorption technology. Interlinking such biotechnologies is potentially possible in future applications to reduce the cost of the biosorption technology and duplicate the benefits of biological H₂ production technology.     

Hydrogen production from C1 compounds by a novel marine hyperthermophilic archaeon Thermococcus onnurineus NA1

A novel marine hyperthermophile, Thermococcus onnurineus NA1, was found to grow on C1 carbon compounds, such as formate and carbon monoxide (CO), and produce hydrogen (H₂). In the present study, the growth and H₂ production of NA1 were examined to determine its potential as H₂ producer. NA1 showed relatively high specific growth rates, 0.48 h⁻¹ and 0.40 h⁻¹ with CO (20%, v/v) and formate (100 mM), respectively, when cultivated in batch mode in a minimal salt medium fortified with 1.0 g L⁻¹ yeast extract. On the other hand, cell growth in both cases stopped at approximately 6 h and the final cell densities were extremely low at 18.2 and 12.1 mg protein L⁻¹ with CO and formate, respectively. The maximum final cell density could be improved greatly to 36.0 mg protein L⁻¹ by optimizing CO content (50%, v/v) and yeast extract concentration (4.0 g L⁻¹), but it was still very low. During the cell growth, formate and CO were used as energy source rather than carbon source. In the resting cell experiments, NA1 exhibited remarkably high H₂ production activities as 385.0 and 207.5 μmol mg protein⁻¹ h⁻¹ for CO and formate, respectively. When formate (100 mM) or CO (100%, v/v) was added repeatedly at 30–35 h intervals, NA1 showed consistent H₂ production for 3 cycles with a yield of approximately 1.0 mol H₂ mol⁻¹ for both CO and formate. This study suggests that T. onnurineus NA1 has a high H₂ production potential from formate or CO but a method for achieving a high cell density culture is needed.     

Optimization of molecular hydrogen production by Rhodobacter sphaeroides O.U.001 in the annular photobioreactor using response surface methodology

Photofermentative H₂ production at higher rate is desired to make H₂ viable as cheap energy carrier. The process is influenced by C/N composition, pH levels, temperature, light intensity etc. In this study, Rhodobacter sphaeroides strain O.U 001 was used in the annular photobioreactor with working volume 1 L, initial pH of 6.7 ± 0.2, inoculum age 36 h, inoculum volume 10% (v/v), 250 rpm stirring and light intensity of 15 ± 1.1 W m⁻². The effect of parameters, i.e. variation in concentration of DL malic acid, L glutamic acid and temperature on the H₂ production was noted using three factor three level full factorial designs. Surface and contour plots of the regression models revealed optimum H₂ production rate of 7.97 mL H₂ L⁻¹ h⁻¹ at 32 °C with 2.012 g L⁻¹ DL malic acid and 0.297 g L⁻¹ L glutamic acid, which showed an excellent correlation (99.36%) with experimental H₂ production rate of 7.92 mL H₂ L⁻¹ h⁻¹.     

Hydrogen production from soft-drink wastewater in an upflow anaerobic packed-bed reactor

The production of hydrogen from soft-drink wastewater in two upflow anaerobic packed-bed reactors was evaluated. The results show that soft-drink wastewater is a good source for hydrogen generation. Data from both reactors indicate that the reactor without medium containing macro- and micronutrients (R2) provided a higher hydrogen yield (3.5 mol H₂ mol⁻¹ of sucrose) as compared to the reactor (R1) with a nutrient-containing medium (3.3 mol H₂ mol⁻¹ of sucrose). Reactor R2 continuously produced hydrogen, whereas reactor R1 exhibited a short period of production and produced lower amounts of hydrogen. Better hydrogen production rates and percentages of biogas were also observed for reactor R2, which produced 0.4 L h⁻¹ L⁻¹ and 15.8% of H₂, compared to reactor R1, which produced 0.2 L h⁻¹ L⁻¹ and 2.6% of H₂. The difference in performance between the reactors was likely due to changes in the metabolic pathway for hydrogen production and decreases in bed porosity as a result of excessive biomass growth in reactor R1. Molecular biological analyses of samples from reactors R1 and R2 indicated the presence of several microorganisms, including Clostridium (91% similarity), Enterobacter (93% similarity) and Klebsiella (97% similarity).