Optimization of Thermodynamic Parameters of the Biological Hydrogen Methanation in a Trickle-Bed Reactor for the Conditioning of Biogas to Biomethane

1. Introduction

Biogas is a natural secondary energy source produced by the microbial degradation of biomass under anaerobic conditions, the combustion of which does not lead to an accumulation of CO₂ in the atmosphere. Biogas consists of 50 - 75 % CH₄ and about 25 - 50 % CO₂, less than 10 % water vapor, and trace amounts of H₂ (< 1 %) and hydrogen sulfide (< 3 %) [1], which corresponds to a calorific value of the gas at standard temperature and pressure of 17.95 to 25.12 MJ/m³ [2]. Currently, desulphurized and dried biogas is mostly used as a fuel in combined heat and power units (CHP) in Germany to generate electrical and thermal energy [3]. The role of biogas in power generation and distribution can be significantly expanded if the quality of biogas is brought up to the standards of natural gas fed into to the gas grid. The natural gas grid in Germany is divided into low-calorific gas (L-gas) with an energy content of 28.8 to 32.4 MJ/m³ and high calorific gas (H-gas) with an energy content of 36.0 – 43.2 MJ/m³. To be fed into the natural gas grid, the calorific value of the biogas must first be raised – which is usually done by removing CO₂. At a CH₄ content of 95% - the threshold value for feeding into the H-gas grid biogas reaches an energy content of 34.1 MJ/m³ [4]. In a second step, the calorific value of this biomethane has to be adjusted to the calorific value of the natural gas in the gas pipeline, usually by adding propane or butane. There are already several CO₂ removal technologies on the market: amine scrubbing, pressure swing absorption, water scrubbing, organic physical scrubbing, cryogenic distillation, and membrane separation [5]. Another promising approach for biomethane production is biological hydrogen methanation (BHM). Following the concept of power-to-gas (PtG), the excess electrical energy from renewable energy sources is used in an electrolyzer that produces “green” H₂. It is then fed into a trickling-bed reactor (TBR) together with CO₂, which is an integral part of a biogas mixture. In the TBR, hydrogenotrophic microorganisms immobilized on plastic carrier bodies will convert both gases into CH₄. Previously conducted studies on CO₂-methanation have shown the influence of operating parameters such as pressure, temperature, and drip interval on process performance in TBRs [6,7,8,9,10]. Researchers found, that when the pressure was increased from 1.5 to 9 bar, the CH₄ content increased simultaneously at mesophilic temperatures of 41 °C [6]. The effect of temperature was investigated by [7] using the same experimental setup. It showed that the conversion rates of H₂ and CO₂ increase, leading to an increasing CH₄ content at increasing temperatures from 40 to 55 °C. The analysis of [10] and [9] achieved CH₄ concentrations above 98 % at ambient pressure and argued the H₂:CO₂ ratio, pH control, and sufficient nutrient supply as limiting factors. At the same time, it was shown in [9] that CH₄ concentrations above 90 % can be achieved in TBR with established microbial culture under thermophilic conditions (55 °C) and ambient pressure at H₂ gas feed rate above 23.2 m³/(m³·d).

Since previous studies focused mainly on CO₂-methanation, which was later referred to as mono-methanation, this study analyzes the co-methanation of biogas and H₂ in a TBR. The objective of the study was to verify the possibility of upgrading biogas with 55 % CH₄ to biomethane with 95 % CH₄ with BHM, and to optimize the thermodynamic parameters of the operating process by applying a design of experiments (DoE).

2. Materials and Methods

2.1. Experimental Setup

The experiments were carried out in the laboratory of the State Institute of Agricultural Engineering and Bioenergy at the University of Hohenheim. The methanation plant was described in detail in [6] and a simplified process schematic of the experimental plant is shown in Figure 1. The plant consists of identical TBRs sharing a single circulation pump that moves the process liquid from the bottom of the reactor to the sprayer at the top. Instead of the mono-methanation of CO₂ and H₂, co-methanation of a standard biogas mixture consisting of 45 % CO₂ and 55 % CH₄ together with H₂ was used (Quality 3.0, Westfalen AG, Germany). The incoming gases were fed to the reactors via mass flow controllers (MFC Type 8742, Buerkert, France) at flow rates of 11.25 L/h for H₂ and 6.25 L/h for the biogas, which met the stoichiometric ratio of 4:1 according to the Sabatier reaction. The flow rates of the educts were adjusted so that the total hourly flow rate of the reacting gases H₂ (11.25 L/h) and CO₂ (2.81 L/h) is comparable to the total volume of the reactors (14.5 L), which consist of 13 L of gaseous main body and 1.5 L of liquid sump together with the periphery. In this case, the gas production at varying operating parameters becomes clear, since the measurement of the product gas quality was performed once per hour and per reactor. During the experiments, the main thermodynamic parameters, namely temperature and pressure, were varied in a certain range to find the optimal operating point.

As a feed liquid, the effluent from the methanogenic stage of a two-stage anaerobic digestion (AD) plant, as described in [11], was introduced into the periphery of the experimental plant to regularly trickle onto the plastic supports with immobilized biofilm inside the methanation column. According to [12], the composition of nutrients required for the BHM is similar to that required for the AD process, allowing liquid transfer between the two plants. Beyond the microbial nutrient removal, the process liquid was constantly diluted by water formation in accordance with the Sabatier reaction. Thus, at the beginning of each experimental phase, the liquid collected in the sump (about 1.3 L) – the lowest point of the experimental plant – was replaced by the fresh feed liquid containing planktonic microorganisms, nutrients and trace elements [13]. Trickling of the carrier bodies occurred once per hour for three minutes.

The product gases leaving the reactors were cooled to remove water vapor. In addition, helium gas (Quality 3.0, Westfalen AG, Germany) was supplied as a tracer gas at a flow rate of 1 L/h and the percentage of each gas component in the gas mixture was measured using a gas chromatograph (micro-GC FUSION Gas Analyzer (Inficon, USA)). Measurements were performed continuously, once per hour for each reactor, resulting in 24 gas quality measurements per day per reactor. Based on the measured content of the tracer gas supplied at a known flow rate, it was possible to calculate the flow rates of all gas components (H₂, CO₂, CH₄, O₂, N₂, H₂S) potentially present in the produced gas mixture.

2.2. Design of Experiment

A DoE was prepared for the study of co-methanation of biogas and H₂ to optimize the thermodynamic parameters using JMP Pro software (SAS Institute, USA). Since [7] recommended experiments at higher temperatures, a temperature in the range of 50 to 60 °C was set, as first predictor variable. Second, a pressure in the range of 2 to 9 bar was chosen. The maximum values are limited by the reactor design. Therefore, the optimal production point can be estimated using a second-order response surface model. Since methanogens are slow to adapt to a changing temperature [14,15], the parameter was set as a “hard-to-change factor”, resulting in a split-plot design, that can be seen in the temperature blocks in Table 1. The different series of experiments are abbreviated as T and the corresponding temperature and P with the corresponding pressure. The minimum number of runs for the given specifications in the custom design is six. However, to minimize the probability of error, the number of runs was set to the number of possible combinations: $3^2=9$. I-optimality was chosen as the optimality criterion, which aims to minimize the variance of the predictions over the relevant range of predictor variables, offering significant advantages in terms of improved prediction compared to the commonly used D-optimality criterion [16]. The odd pressure in the sixth run is dictated by the experimental design, which attempts to minimize prediction errors.

The duration of each experimental phase was set at 144 hours (six complete days), including an adjustment period of 48 hours. After 48 hours, pH and CH₄ production were stabilized, and the above measurements were considered for further analysis.

2.3. Analytical

During each experimental phase, samples of the process liquid were taken three times and further analyzed in the local laboratory of the state institute. The samples were analyzed for volatile fatty acids (VFA) content (acetic acid, n-butyric acid, iso-butyric acid, propionic acid, n-valeric acid, iso-valeric acid, and caproic acid) in the analytical laboratory to exclude or confirm the activity of acetoclastic microorganisms in the methanation process. At the same time, the measurement of chemical oxygen demand (COD) was carried out to determine the percentage of organic degradable material. In practice, the COD of the BHM effluent should ideally be as low as possible, so that it can be discharged into nature without any problems. Total carbon (TC) and total organic carbon (TOC) analysis of the process fluid was performed to balance the amount of carbon involved in the BHM reaction.

VFA concentration was measured using gas chromatography (GC 2100Plus, Shimadzu with an FID-detector and a capillary column WCOT Fused Silica, Varian, Palo Alto USA). COD content was measured using a sensor array photometer (Hach Lange Type LASA 20). A TOC/TNb analyzer (Analytik Jena AG Type multi N/C®, Jena, Germany) was used to measure TC and TOC.

The pH was measured separately in the sump of each reactor using combined pH and Redox sensors (Endress and Hauser AG, Reinach, Switzerland). Pressure in each reactor was measured using an absolute pressure transmitter (ABB Ltd., Minden, Germany). Temperature was determined using compact thermometers (Endress and Hauser AG, Reinach, Switzerland). All sensor data were recorded once per minute and logged in the database.

2.4. Calculations

Both analog sensor data and gas quality measurements were evaluated for the analysis. Since no significant difference in reactor performance was found, average and median values of the process parameters are reported in this study. To evaluate the performance of the reactors, the following key values were determined based on the collected data:

• Methane formation rate (MFR);
• Gas hourly space velocity (GHSV);
• Retention time (RT);
• and the conversion rates of H2 and CO2.

As defined in [17,18], the GHSV represents the ratio between the flow rate of the incoming gases at STP and the volume of the reactor or its catalyst content. In the experimental conditions studied, only the reactant gas components fed into the reactor were considered:

$$GHSV=\frac{F_{C{O}_2}+F_{H_2}}{V_r}[1/h]$$

(1)

Where F_CO2 and F_H2 are the flow rates of CO₂ and H₂ at STP in m³/h and V_r is the reaction volume, V_r = 0.013 m³. RT refers to the time required for the incoming gases to pass through the reactor volume from the inlet at the bottom of the reactor to the outlet at the top [6]:

$$RT=\frac{V_r}{F_{C{O}_2}+F_{H_2}+F_{C{H}_4}}\left[h\right]$$

(2)

Where F_CO2, F_H2, and F_CH4 are the flow rates of CO₂, H₂, and inert CH₄ supplied to the reactor in m³/h at the reactor temperature and pressure [19]. One of the most important parameters related to the performance of the reactor is the MFR, which is determined by the daily flow rate of the product gas and the volume of the reactor [6,19]:

$$MFR=\frac{F_{C{H}_{4_{out}}}-F_{C{H_4}_{in}}}{V_r}[m³/(m³·d\left)\right]$$

(3)

Where F_CH4in is the daily flow rate of CH₄ entering the reactor and F_CH4out is the daily flow rate of CH₄ leaving the reactor in m³/d. Compared to [6], in the present study, the daily flow rate of microbially produced CH₄ was defined as the difference between the measured volumetric flow rate of CH₄ in the product gas and the volumetric flow rate of CH₄ injected into the reactors.

3. Results and discussion

3.1. Operation and performance parameters

The data on the measured parameters as well as the key parameters related to the performance of the reactors during all experimental phases are shown in Table 2. Both temperature and pressure were stable in each experimental phase and did not vary significantly during the experiments. Since the GHSV is related to the fixed parameters of the experimental procedure, its value was kept constant throughout the experiment: GHSV = 1.082 1/h.

The analysis of the flow rates of the incoming and outgoing gases showed some deviation of the H₂:CO₂ ratio from the set stoichiometric value: during all experimental phases there was a slight overshoot of H₂ content. This could be due either to the inaccuracy of the instruments, as reported in [6], or to the fluctuating ratio between CO₂ and CH₄ in the gas mixture supplied to the reactors from the same cylinder. This over-stoichiometric ratio resulted in a better conversion rate of CO₂ compared to the conversion of H₂. On average, the conversion rate of H₂ was 97.124 ± 0.176 % and the conversion rate of CO₂ was 99.941 ± 0.011 % throughout the experiment.

The results of the experiments showed a clear correlation between the quality of biomethane and the H₂:CO₂ ratio (Figure 2). Thus, the purity of the biomethane is determined by the ratio of feed gases, resulting in higher quality the closer the ratio is to 4.

In the experiments where the H₂:CO₂ ratio was 4.123, the content of CH₄ in the product gas was the lowest. This result indicates a threshold value for the ratio of the reaction gases for their successful conversion into biomethane: The H₂:CO₂ ratio should not exceed 4.083 to achieve a biomethane quality of at least 95 %. On the other hand, the conversion of H₂ shows an obvious linear correlation with the purity of biomethane, with higher conversion rates of H₂ leading to higher purity of biomethane (R² = 0.925, see supplementary Materials, Figure S1). At the same time, no obvious correlation was found between the purity of biomethane and the conversion of CO₂ (see supplementary Materials, Figure S2). Since the coefficient of determination of the correlation between the H₂:CO₂ ratio is higher than the H₂ conversion, in practice the dosage of the reaction gases should be controlled and kept as close as possible to the stoichiometric ratio in order to achieve optimal conversion rates of the reactants and maximally reduce the fraction of their residues in the product gas.

The MFR exhibited a relatively constant trend during the experiments confirming the stability of the process and a robust BHM of the injected gases in the presence of inert CH₄ gas. It is worth noting that the MFR increases slightly with increasing pressure and temperature in the reactors, and the obtained results are comparable to the results of [6,8]. At a H₂ gas feed rate of 22.77 m³/(m³·d) (at a flow rate of 11.25 L/h based on the trickle-bed volume V_r = 13 L), the MFR obtained in our study was comparable to the results of [9] for a similar H₂ feed rate. However, the median CH₄ concentration at a temperature of 55 °C was sufficiently higher in our experiments than in [9]: it was above 94 % for all experimental setups and reached 95.494 % at a pressure of 4.275 bar. These results underline the importance of pressure for biogas upgrading in TBRs. Overall, the increase in MFR and the percentage of bacterially converted methane CH_4conv demonstrate the higher bacterial activity with increasing temperature and pressure, with the temperature effect being more significant than the pressure increase (see supplementary Materials, Figure S3).

Regarding the quality of the upgraded biogas, the highest biomethane purity was obtained at a pressure of 4.275 bar and a temperature of 55 °C, and at a pressure of 9 bar and a temperature of 50 °C (Figure 3). In all experimental setups the lowest biomethane quality for “low standard” gas grid is achieved, and in the two cases mentioned, the median value of the quality for “high standard” gas grid is achieved.

3.2. Optimizing operating parameters

To evaluate the optimal operating parameters in JMP Pro, three key process performance values were added to the custom design. Biomethane quality, H₂ conversion and MFR were set as equally weighted variables. The standard least square model is chosen to analyze the DoE and the desirability function is optimized over all responses. The maximum desirability is obtained at 50 °C and 9 bar. However, Figure 4 shows no maximum of a key value in the specified intervals of temperature and pressure. The prediction formula shows a negative correlation between biomethane quality, H₂ conversion and increasing temperature, which contradicts the conclusion of [7]. On the other hand, there is a positive correlation between increasing pressure and both variables. As for MFR, both temperature and pressure show a parabolic correlation with this performance parameter. This indicates that further experiments with higher pressure and wider temperature interval are needed to validate the prediction formula found. Furthermore, when the fluctuating ratio of H₂:CO₂ in the reactants is included in the prediction formula, an effect on the biomethane quality and H₂ conversion rate, but not on the MFR can be seen. This shifts the optimal process temperature to 60 °C, which underlines the previous conclusion to use a wider temperature interval in further experiments. On the other hand, ensuring a constant H₂:CO₂ ratio must be considered when planning future experiments.

3.3. Analysis of the process liquid

Laboratory analyses of the process liquid show an increase in pH with increasing temperature and pressure (Figure 5 (a)). These results are in contrast to the results of [6], where the pH decreased with increased pressure due to the formation of carbonic acid. This may be due to the amount of inert CH₄ in the reactor, which leads to a lower partial pressure of CO₂. The result of [20] also show the importance of pH for process stability, including microbial growth. Since the solubility of CO₂ increases not only with pressure but also with increasing pH, an over stochiometric feed could lead to process disturbances and inhibition of methanogenesis [21].

Since the formation of VFAs occurred in all reactors, the activity of acetoclastic methanogens could not be excluded. The predominant acids found in the reactors were acetic and propionic acids (Figure 5 (b)). In addition, both COD and TOC analyses show no significant difference between the effluents from the reactors and the feed liquid (Figure 6 (a) and (b)). The TC and TN analyses also showed that the inorganic carbon and TN content in the reactor effluent is significantly lower compared to the feed liquid (Figure 6 (c) and (d)). The depletion of both nutrients from the process liquid over time may indicate biomass growth, as carbon serves as an energy source and nitrogen compounds are required for protein synthesis [22]. At the same time, the relatively low C/N ratio (below 30, as mentioned by [22]) could also be related to VFA formation within the reactors.

5. Conclusions and outlook

The present study shows that biogas can be upgraded to biomethane quality by using BHM. By applying a DoE, the optimum process parameters in terms of pressure and temperature were determined with the given experimental setup. At 50 °C and 9 bar, the MFR was 5.295 ± 0.216 m³/(m³·d), CH₄ content was 95.614 ± 0.151 % and H₂ conversion was 97.535 ± 0.166 %. However, no maximum value was obtained in the prediction formula for the key values, suggesting that further experiments with longer intervals are needed. In addition, a quality of over 93 % CH₄ in the product gas was achieved in all experimental runs, which is sufficient for injection into the “low” gas grid. However, under real industrial conditions, the ratio between CO₂ and CH₄ in the biogas may vary, depending on the type of substrates fed or the conditions in the digester. In order to maintain the quality of biomethane fed into the gas grid, the H₂:CO₂ ratio in the educts should be controlled. According to the results of the present study, the H₂:CO₂ ratio should not exceed 4.083:1. At the same time, in the works of [6,9,10] for mono-methanation of CO₂, the lower limits for this parameter were 3.75:1. Therefore, an optimal H₂:CO₂ ratio must be ensured in practice. In addition, an economic analysis weighing the benefits of higher temperature and pressure against manufacturing and operating costs could improve process optimization.