Abstract

Within the European circular economy roadmap, it is important for wastewater treatment plant (WWTPs) to recover energy and become energy-neutral or -positive. In the last few years, it has become increasingly interesting to boost energy recovery through the biogas upgrading. The aim of this work is to study a rapid hydrogenotrophic methanogenic culture enrichment strategy capable of limiting the organic degradation unbalance and allowing a fast start-up phase of the in situ biogas upgrading reactors, at pilot- or full-scale. The approach was tested with two, plus one control, laboratory-scale continuous stirred tank reactors filled with anaerobic sludge collected from a full-scale WWTP. The experimentation lasted 50 days and was divided into five phases: the anaerobic digestion start-up followed by four H2 injection phases (H2/CO2 ranging from 1:1 to 4:1 on molar basis). Despite a temporary slight increase in the total concentration of volatile fatty acids during phase II (2.56 gCH3COOH·L−1), and in phase III a mild pH increase (anyway, below 7.4) indicating the expected CO2 depletion, the strategy proposed was effective. In the last phase, in the biogas a methane content of about 80% was achieved, thus suggesting that the use of H2/CO2 above the stoichiometric value could further improve the biological biogas upgrading.

INTRODUCTION

Within the circular economy context, it is important for wastewater treatment plant (WWTPs) to reduce the energy demand and move towards energy self-sufficiency. According to Silvestre et al. (2015), an energy amount of about 3.2 kJ·gTS−1 (TS: total solids) is contained in raw wastewater, and an average amount of energy of 0.35 kJ·gTS−1 is required for sewage treatment; if captured and managed efficiently, sludge generated in WWTPs could yield substantial energy in the form of biogas, potentially turning a WWTP into a net energy producer rather than a consumer. However, currently this is still far from being viable (Shen et al. 2015).

Anaerobic digestion (AD) is the technology commonly used to recover energy from organic streams. Biogas produced from AD of primary and secondary sludge from WWTPs consists mainly of methane (55–70%) and carbon dioxide (30–45%). Besides CH4 and CO2, raw biogas also contains small amounts of nitrogen (0–15%), oxygen (0–3%), water (1–5%), hydrocarbons (0–200 mg·m−3), hydrogen sulphide (0–10,000 ppmv), ammonia (0–100 ppmv), and siloxanes (0–41 mgSi·m−3) (Sun et al. 2015; Awe et al. 2017).

The lower heating value of biogas is usually found to be roughly around 23,400 kJ·Nm−3 depending on methane percentage (Silvestre et al. 2015). Biogas upgrading relies on the contaminant's removal or transformation from the raw biogas, in order to produce a final output gas consisting of higher methane concentration. Typically, the removal of moisture, H2S and CO2 represents the most important upgrading steps (Miltner et al. 2017). If the upgraded biogas is purified to natural gas standards, then the final gas product is called biomethane (Kougias et al. 2017).

Different technologies are currently used for biogas cleaning and upgrading. Physical (condensation) and chemical (adsorption or absorption) drying methods are used to remove water. Two procedures commonly used to remove H2S during digestion are air/oxygen dosing to the biogas, and the addition of Fe2+ or Fe3+ in the form of FeCl2, FeCl3 and FeSO4 into the digester or to the organic feed. Techniques such as adsorption on iron oxide or hydroxide or activated carbon, absorption with gas–liquid contactors (spray or packed bed towers using water, organic solvents or aqueous chemical solutions with H2S conversion to elemental sulfur or metal sulfide) and membrane separation are commonly adopted to remove H2S after digestion (Muñoz et al. 2015). Subsequently, trace components like siloxanes, hydrocarbons, ammonia, oxygen, carbon monoxide, and nitrogen can require extra removal steps, if not sufficiently removed by other treatment steps. Finally, the bulk CO2 content must be separated from CH4. Several commercial technologies are currently available: pressure or vacuum swing adsorption, membrane separation, physical or chemical CO2 absorption (scrubbing with water, CO2-reactive absorbents or organic solvents) and cryogenic separation (Ryckebosch et al. 2011; Awe et al. 2017).

Nowadays, those commercial CO2 separation techniques are facing significant challenges in terms of energy/chemicals consumption and operating costs, making the upgraded gas expensive and not always affordable from an economic point of view. An alternative solution currently attracting many researchers is the biological biogas upgrading via the enhancement of hydrogenotrophic methanogenesis (Kougias et al. 2017).

The methane formation, known as methanogenesis, is the final AD process step and it is exclusively carried out by methanogenic members belonging to the Archaea domain. Organic substrates are converted to methane by distinct but concomitant methanogenic pathways operative in phylogenetically diverse methanogens: the acetoclastic methanogens, which convert acetate into CH4 according to Equation (1), and the hydrogenotrophic methanogens, which convert H2 and CO2 into CH4 without other organic carbon sources, according to Equation (2) (Kern et al. 2016).  
formula
(1)
 
formula
(2)
Although acetoclastic methanogens have a major role in CH4 production (approximately 70%), methane can be also produced from hydrogen plus carbon dioxide or formate (Smith & Mah 1966). At high H2 concentrations (e.g. >500 Pa), acetogenesis or methanogenesis from H2 + CO2 is favored, and at low concentrations (e.g. <40 Pa), oxidation of the acetate occurs (Demirel & Scherer 2008). Thus, at low H2 partial pressure (i.e. the normal anaerobic process), hydrogenotrophic methanogens maintain low H2 partial pressure necessary for the growth of intermediate syntrophic bacteria (Zinder 1994).

In the last few years, three applications which rely on the enhancement of hydrogenotrophic methanogens for biogas upgrading have been mainly studied: in situ, ex situ (Luo & Angelidaki 2013; Kougias et al. 2017) and more recently a hybrid system, which couples the in situ and the ex situ in one operational unit, to benefit from both system advantages (Corbellini et al. 2018).

The hydrogenotrophic biogas upgrading process has several advantages mainly related to the higher CH4 final volume and to CO2 removal from biogas, which would decrease the costs for the upgrading of biogas to natural gas quality. However, in order to convert the major part of CO2, H2 has to be generated by an external source. Recent literature has reported on the use of bio-electrochemical systems, already applied for nutrient, metal and energy recovery as well as for wastewater treatment, coupled with AD in order to enhance CH4 production while removing CO2 in biogas. Specifically, bio-electrochemically assisted AD (AD-BEC) consists of applying, a relatively low, external potential to a conventional anaerobic digester, then making possible the simultaneous biogas production and upgrading; hydrogenotrophic methanogenesis and electro-methanogenesis are the two main processes through which CO2 and electrons from the cathode electrode are directly used for CH4 production AD-BEC systems (Dou et al. 2018). The other sustainable technology to produce H2 is water electrolysis utilizing excess energy from windmills or solar power stations (Ullah Khan et al. 2017); in this respect, biological upgrading represents a highly promising approach to connect net electricity to the natural gas grid via water electrolysis (Lecker et al. 2017).

Moreover, as for the in situ pathway, the utilization of the existing infrastructure of biogas plants and the need for lower technical requirements would result in reduced operational and investment cost and energy compared to available technologies. Finally, possible unconverted hydrogen mixed with methane would improve the combustion properties of biogas as fuel (5–30% hydrogen by volume) (Luo et al. 2012).

Focusing on the in situ application, three main issues have been identified: (1) the low solubility of H2; (2) the addition of hydrogen to a biogas reactor might cause problems or even a breakdown of the process: the increase of hydrogen partial pressure can lead to a subsequent inhibition of volatile fatty acids (VFA) degradation (propionate and butyrate); (3) H2 injection exceeding the 4:1 stoichiometric ratio between CO2 and H2 could result in CO2 depletion, and thus lead to an increase of pH: too alkaline pH values may limit the methanogenic activity, while a depletion of CO2 could entail a substrate inhibition for autotrophic hydrogenotrophic methanogens, which rely on CO2 as a carbon source (Luo et al. 2012; Rachbauer et al. 2016). Luo & Angelidaki (2013) tried to overcome the pH increase by means of co-digestion of cattle manure with an acidic substrate such as cheese whey. However, since co-digestion cannot always be adopted especially within the field of municipal wastewater treatment, other optimization modes need to be identified. Few studies have focused on the development of efficient hydrogenotrophic methanogen enrichment in situ to overcome the risk of total volatile fatty acids (TVFA) shock caused by the unadapted consortia at high H2 concentration. Recently, Agneessens et al. (2017) tested H2 pulse injections in order to induce modulation of the microbial community, resulting in an increased H2 uptake. Xu et al. (2015) performed a continuous cultivation in an up-flow anaerobic sludge blanket (UASB) reactor with H2/CO2 (4:1) as the sole substrate.

The aim of this work was to develop a fast hydrogenotrophic methanogenic culture enrichment strategy capable of limiting organic biodegradation unbalance and to allow a fast start-up of in situ biogas upgrading reactors, at the pilot- or full-scale, at mesophilic conditions. The effectiveness of the enrichment procedure was evaluated in terms of methane content and specific methane production in the output gas, H2 conversion efficiency, pH trends, VFA concentrations and speciations dynamics, chemical oxygen demand (COD) mass-balance and by specific hydrogenotrophic methanogens activity (SHMA) measurement.

MATERIAL AND METHODS

Semi-continuous experiment: reactor set-up and operation

The hydrogenotrophic methanogenic culture enrichment was performed using three continuous stirred tank reactors (CSTRs) (total volume = 2.4 L; working volume = 1 L), namely R1 as control (no hydrogen injection), and R2 and R3 as two replicates; the experimental set-up is shown in Figure 1. All reactors were incubated at 35 °C and continuously mixed at 150 rpm, by means of a magnetic stirrer, in order to maximize H2 dissolution into the liquid phase.

Figure 1

Graphical representation of the experimental set-up. Reactor R1, on the left, was run as a control reactor; reactors R2 and R3, on the right, were run as replicates of the hydrogenotrophic methanogenic enrichment in situ test (GC: gas chromatography).

Figure 1

Graphical representation of the experimental set-up. Reactor R1, on the left, was run as a control reactor; reactors R2 and R3, on the right, were run as replicates of the hydrogenotrophic methanogenic enrichment in situ test (GC: gas chromatography).

A mixture of primary and biological sludge, collected from a full-scale municipal WWTP (Bresso – Seveso Sud, Milan, Italy), was manually fed in semi-continuous mode (5 days per week). A dose of 70 mL of fresh sludge mixture was adopted, corresponding to an organic loading rate (OLR) of 1 gVS·L−1·d−1 (VS: volatile solids), and to a hydraulic retention time (HRT) of 15 days. Main characteristics of substrate and inoculum used in the tests are summarized in Table 1.

Table 1

Average characteristics of inoculum and feeding substrate used. The sludge mixture was composed of primary and biological sludge

Parameters Unit Sludge mixture Inoculum 
Total solids (TS) g/kg 22.4 ± 3(1) 24.9 ± 3 
Volatile solids (VS) g/kg 14.7 ± 2 14.7 ± 3 
VS/TS 66 ± 0 59 ± 1 
TKN mgN/kg 749 1,350 
COD g/kg 10 5.5 
TVFA mgCH3COOH/L 824 255 
Parameters Unit Sludge mixture Inoculum 
Total solids (TS) g/kg 22.4 ± 3(1) 24.9 ± 3 
Volatile solids (VS) g/kg 14.7 ± 2 14.7 ± 3 
VS/TS 66 ± 0 59 ± 1 
TKN mgN/kg 749 1,350 
COD g/kg 10 5.5 
TVFA mgCH3COOH/L 824 255 

(1) ± standard deviation.

The three reactors were inoculated with the digestate taken from the full-scale digester in the same WWTP (Bresso – Seveso Sud, Milan, Italy) where the sludge mixture used as feeding was collected. Bottles, stored at mesophilic conditions (35 °C), were then flushed with nitrogen gas (N2) in order to ensure anaerobic conditions. A mineral medium solution, containing macro- and micro-nutrients, was added to the three anaerobic reactors in the ratio 1:10 with respect to the working volume, in order to avoid lack of trace elements during culture enrichment. The mineral medium was prepared according to Angelidaki et al. (2009).

The enrichment test on reactors R2 and R3 was divided into five phases; the operating conditions of each of them are shown in Table 2. The start-up phase (I), aimed at the acclimation of the biomass, lasted 19 days; during phase I, no hydrogen was added–only the sludge mixture; then, four enrichment phases (from II to V), lasting a total of 29 days, were implemented: during phases II to V, besides the feeding of the sludge mixture, H2 was dosed at different and increasing H2/CO2 ratios. Specifically, the H2/CO2 ratio was raised from 1:1mol H2/mol CO2 to the stoichiometric value of 4:1mol H2/mol CO2. Before the daily addition of the sludge mixture by means of a syringe, gas volume and composition were measured and reactors were vented to the atmospheric pressure. The volume of hydrogen to be dosed was then calculated based on the average daily flow rate of CO2 produced during the previous phase; thus, H2 was injected using a gas-tight syringe. Table 2 summarizes the operative conditions that were adopted for the enrichment reactors (R2 and R3); the OLR is reported with reference to the contribution given by the sludge mixture only (OLRSM), and to the total COD fed including both the hydrogen (8g COD/g H2) and the sludge mixture (OLRtot).

Table 2

Operational parameters adopted during the five experimental phases of the enrichment trial

Experimental phase Duration OLRSM OLRtot H2/CO2 ratio 
(days) (gCOD·L1·d1(gCOD·L1·d1(mol H2/mol CO2
Start-up 19 – 
II Enrichment 1.05 1:1 
III Enrichment 1.07 2:1 
IV Enrichment 1.09 3:1 
Enrichment 1.12 4:1 
Experimental phase Duration OLRSM OLRtot H2/CO2 ratio 
(days) (gCOD·L1·d1(gCOD·L1·d1(mol H2/mol CO2
Start-up 19 – 
II Enrichment 1.05 1:1 
III Enrichment 1.07 2:1 
IV Enrichment 1.09 3:1 
Enrichment 1.12 4:1 

Monitoring of the process and analytical methods

Reactors were fed five times per week: TS, VS and COD were measured in the feeding sludge mixture according to Standard Methods 2540 for solids and 5220 for COD (APHA 2005). Total Kjeldahl nitrogen (TKN) was also measured according to the ISO 5663:1984 (ISO 1984).

Corresponding to each feeding, digestate was discharged and analysed for TS, VS, pH and VFA. The pH, which was not controlled during the experiment in order to follow its variation over time, thus simulating real conditions, was directly measured in samples by means of a portable multi-probe meter (Hach-Lange, HQ40D). The VFA (acetic, propionic, isobutyric, butyric, isovaleric and valeric) concentrations were determined according to Standard Method 5560 (APHA 2005), using a gas chromatograph (DANI Master GC) coupled with a flame ionization detector (FID Nukol fused silica).

The manometric method was used for monitoring the biogas production. The pressure was daily measured using a digital manometer (Keller LEO 2) by puncturing the rubber septum; the volume of gas produced during the anaerobic degradation was computed from pressure data, according to the ideal gas law. Biogas composition (CO2, CH4, H2, O2, N2) was analysed three times per week by using a gas chromatograph (DANI Master GC Analyser equipped with two columns, HayeSep Q and Molesieve 5A).

The volume of hydrogen (Dose(H2)phase_i) to be daily dosed in reactors R2 and R3 during the ith enrichment phase (from II to V) was expressed as mLH2·d−1 and calculated according to Equation (3):  
formula
(3)
where (H2/CO2)phase_i is the hydrogen to carbon dioxide molar ratio to be used in phase i, and Q(CO2)phase_i is the rate of carbon dioxide daily produced and released in biogas during the previous phase ‘i 1’ and expressed in mLCO2·d−1.
Two coefficients, both expressed as percentages, the CO2 conversion efficiency and the H2 utilization efficiency, were also evaluated in order to monitor the enrichment evolution. CO2 conversion efficiency (%) of the ith enrichment phase was derived according to Equation (4):  
formula
(4)
where (CO2)phase_I is the rate of CO2 daily produced at the stable point of phase I (mLCO2·d−1), and (CO2)phase_i is the average rate of carbon dioxide produced during phase i. The H2 utilization efficiency was calculated according to Equation (5):  
formula
(5)
where (H2)phase_i is the daily hydrogen amount measured in the gas phase (mLH2·d−1).

Furthermore, the COD mass-balance was evaluated for both reactors. Influent COD was calculated considering the COD of both the sludge mixture and the H2 injected, while the effluent COD was the sum of the COD discharged in both the liquid and the gas phases, including: the non-degradable VS, multiplied by a conversion factor of 0.7 gCOD·gVS−1 based on the substrate characteristics; for methane, 0.35 NmLCH4/gCOD was used; and the unconverted H2 multiplied by a factor of 0.7 mgCOD/NmLH2.

SHMA procedure

The SHMA of the biomass was assessed both at the beginning of the experimentation and at the end of each enrichment phase. SHMA was assessed at mesophilic conditions (35 ± 0.5 °C) adopting an internal protocol based on the manometric method. In detail, the digestate taken from the two reactors was tested in serum vials of 40 mL, fluxed with nitrogen gas and with a gas mixture composed of H2 and CO2 in the stoichiometric ratio of 4:1 (molar base). Tests were run for 7 hours, at mesophilic conditions and adopting a ratio between the headspace volume and the liquid volume of 2. H2 diffusion in the liquid phase was ensured by means of a magnetic stirrer set at 150 rpm. Headspace pressure was measured every 20 minutes, and methane production was derived assuming that one mole of H2 produces 0.25 moles of CH4; the SHMA expressed as NmLCH4·gVS−1·h−1 was calculated as follows (Equation (6)):  
formula
(6)
where X (gVSL·L−1) is the amount of VS of the digestate dosed in each bottle, and the term dV(CH4)/dt (NmLCH4·h−1) refers to the maximum slope of the cumulative methane production trend over time.

Statistical analysis

Statistical analysis was carried out using the software SPSS v.25 aimed at statistically assessing: (i) the significance of the observed differences of reactors R2 and R3 compared with the control reactor R1; (ii) the reproducibility of the two replicates, reactors R2 and R3. Since variables were not normally distributed, the nonparametric Mann–Whitney U-test (significance level = 0.05) was used to compare the dependent variables (methane content in the biogas and biogas production rate) for two independent groups (R1 and R2, R1 and R3, R2 and R3).

RESULTS AND DISCUSSION

Performance of the semi-continuous reactors

Table 3 summarizes the results obtained during the five experimental phases in the three reactors: analyses on biogas and on discharged effluents are shown, as well as performance parameters.

Table 3

Characteristics of biogas and digestate during the experimental phases I–IV, and performance parameters measured on reactors R1 (control), R2 and R3 (enrichment reactors)

  I (pre-H2)
 
II (H2/CO2 1:1)
 
III (H2/CO2 2:1)
 
IV (H2/CO2 3:1)
 
V (H2/CO2 4:1)
 
  R1 R2 R3 R1 R2 R3 R1 R2 R3 R1 R2 R3 R1 R2 R3 
H2 flow rate (NmL·d−1NA* NA* NA NA* 62 62 NA* 102 98 NA* 133 124 NA* 222 204 
Biogas production rate (NmL·d−1217 ± 20 205 ± 9 213 ± 18 215 ± 20 218 ± 26 197 ± 7 226 ± 31 262 ± 4 209 ± 17 234 ± 5 253 ± 16 253 ± 30 233 ± 30 250 ± 17 Nd** 
Biogas composition 
  • – CH4 (%)

 
72.6 ± 1.1 73.0 ± 1.0 72.4 ± 0.5 72.2 ± 0.4 73.5 ± 0.2 74.5 ± 0.3 72.5 ± 0.3 75.2 ± 0.3 76.3 ± 0.3 71.3 ± 0.9 77.2 ± 0.3 77 ± 0.5 71.3 ± 0.5 80.2 ± 0.9 75.31 ± 1.1 
  • – CO2 (%)

 
27.5 ± 0.7 27.0 ± 1.0 27.6 ± 0.6 26.3 ± 0.5 26.5 ± 0.2 25.4 ± 0.2 27.5 ± 0.7 24.8 ± 0.6 23.6 ± 0.6 28.6 ± 0.1 22.8 ± 0.1 22.3 ± 0.1 26.3 ± 0.2 20.3 ± 0.2 24.7 ± 0.2 
  • – H2 (%)

 
Na* Na* Na* Na* 2.6 ± 1.4 0.1 Na* Na* Na* 3.6 ± 1.5 
Methane yield (NmLCH4·gVS−1157 149 153 156 160 146 164 169 160 162 194 194 161 195 Nd** 
H2 utilization efficiency (%) Na* Na* Na* Na* 96.0% 99.8% Na* 100% 100% Na* 100% 100% Na* 100% 100% 
CO2 conversion efficiency (%) Na* Na* Na* Na* 41.5% 38.6% Na* 49.1% 26.4% Na* 25.1% 13.4% Na* 45.9% Nd** 
pH (−) 7.3 7.4 7.4 7.2 7.2 7.2 7.2 7.3 7.3 7.0 7.3 7.2 7.1 7.2 7.2 
TVFA (mgCH3COOH·L1834 314 943 632 348 2,560 313 275 1,090 269 412 274 209 184 1,012 
  I (pre-H2)
 
II (H2/CO2 1:1)
 
III (H2/CO2 2:1)
 
IV (H2/CO2 3:1)
 
V (H2/CO2 4:1)
 
  R1 R2 R3 R1 R2 R3 R1 R2 R3 R1 R2 R3 R1 R2 R3 
H2 flow rate (NmL·d−1NA* NA* NA NA* 62 62 NA* 102 98 NA* 133 124 NA* 222 204 
Biogas production rate (NmL·d−1217 ± 20 205 ± 9 213 ± 18 215 ± 20 218 ± 26 197 ± 7 226 ± 31 262 ± 4 209 ± 17 234 ± 5 253 ± 16 253 ± 30 233 ± 30 250 ± 17 Nd** 
Biogas composition 
  • – CH4 (%)

 
72.6 ± 1.1 73.0 ± 1.0 72.4 ± 0.5 72.2 ± 0.4 73.5 ± 0.2 74.5 ± 0.3 72.5 ± 0.3 75.2 ± 0.3 76.3 ± 0.3 71.3 ± 0.9 77.2 ± 0.3 77 ± 0.5 71.3 ± 0.5 80.2 ± 0.9 75.31 ± 1.1 
  • – CO2 (%)

 
27.5 ± 0.7 27.0 ± 1.0 27.6 ± 0.6 26.3 ± 0.5 26.5 ± 0.2 25.4 ± 0.2 27.5 ± 0.7 24.8 ± 0.6 23.6 ± 0.6 28.6 ± 0.1 22.8 ± 0.1 22.3 ± 0.1 26.3 ± 0.2 20.3 ± 0.2 24.7 ± 0.2 
  • – H2 (%)

 
Na* Na* Na* Na* 2.6 ± 1.4 0.1 Na* Na* Na* 3.6 ± 1.5 
Methane yield (NmLCH4·gVS−1157 149 153 156 160 146 164 169 160 162 194 194 161 195 Nd** 
H2 utilization efficiency (%) Na* Na* Na* Na* 96.0% 99.8% Na* 100% 100% Na* 100% 100% Na* 100% 100% 
CO2 conversion efficiency (%) Na* Na* Na* Na* 41.5% 38.6% Na* 49.1% 26.4% Na* 25.1% 13.4% Na* 45.9% Nd** 
pH (−) 7.3 7.4 7.4 7.2 7.2 7.2 7.2 7.3 7.3 7.0 7.3 7.2 7.1 7.2 7.2 
TVFA (mgCH3COOH·L1834 314 943 632 348 2,560 313 275 1,090 269 412 274 209 184 1,012 

*NA, not applicable to this period/reactor; **Nd, not determined due to a technical issue.

In Figure 2, trends over time of the daily biogas volume produced and the daily methane, carbon dioxide, and hydrogen percentages measured are presented for the three reactors. Process stability was monitored by measuring pH and VFA total amount and distribution in order to detect possible accumulation and consequent methanogen inhibitory effect. Figure 3 shows the distribution of VFA measured in the three reactors during all five experimental phases.

Figure 2

Biogas production rate and composition measured for the three reactors during all experimental phases; R1 is the control reactor; R2 and R3 are the two replicates of the enrichment trial.

Figure 2

Biogas production rate and composition measured for the three reactors during all experimental phases; R1 is the control reactor; R2 and R3 are the two replicates of the enrichment trial.

Figure 3

Volatile fatty acids composition during five experimentation phases; R1 is the control reactor; R2 and R3 are the two replicates of the enrichment trial.

Figure 3

Volatile fatty acids composition during five experimentation phases; R1 is the control reactor; R2 and R3 are the two replicates of the enrichment trial.

In general, reproducible results, in terms of biogas composition and methane production, were observed for reactors R2 and R3 until phase IV, thus suggesting that the equipment used was appropriate for the enrichment tests. More significant differences were found in VFA composition, indicating that intermediate compounds in the anaerobic degradation chain may be more affected by slight differences of environmental conditions, between parallel reactors.

The Mann–Whitney U-test, considering all methane content data collected during all five phases, showed that the distribution between R2 and R3 (U = 170, exact significance = 0.204) was the same (mean ranks: R2 = 23.9; R3 = 19.1). Moreover, since biogas production rate data in R3 during phase V were not available, only data collected from phases I to IV were used to test the distribution of biogas production rate. This resulted in the distribution being comparable between R2 and R3 (U = 113, exact significance = 0.195), with mean ranks equal to 20.2 for R2 and 15.7 for R3. Furthermore, the result was the same, displaying only an exact significance value (0.095) closer to the significance level of 0.05, if all biogas production rate data available (phases I to V for R2, and I to IV for R3) were considered.

During phase I, the acclimation of the biomass taken from the full-scale digester to the new operative conditions was ensured; at the end of this phase, indeed, methane production achieved a steady state of about 157 NmLCH4·gVS−1, resulting in a biogas composed of 73% of methane and for the 27% of carbon dioxide. On average, at the end of phase I, pH was slightly lower in the control reactor (7.3) compared to values measured in both R2 and R3 (7.4). During the start-up phase, TVFA were lower for R2 (0.3 g·L−1) compared to both R1 and R3 (approximately 0.7 g·L−1). With reference to TVFA speciation, acetic acid prevailed in R2 (85%), while in reactors R1 and R3 the amount of acetic acid accounted for about 65%, with a content of propionic acid of 15%, indicating that new operative conditions could have led to a slight unbalance on propionate degradation.

Moving to the enrichment phases, during the first period (phase II), a small H2 dose (approximately 62 NmLH2·L−1·d−1) determined different effects on R2 and R3, considering both VFA total composition and speciation and methane yield. Compared to phase I, in the control reactor R1, the specific methane yield remained stable at 156 NmlCH4·gVS−1, while in R2 and R3 it respectively increased and decreased by 5% (160 NmlCH4·gVS−1 and 146 NmlCH4·gVS−1, respectively).

As expected, TVFA concentration in R1 decreased, indicating a complete adaptation of the biomass to the actual operative condition; in contrast, R2 and R3 showed different behaviour. In more detail, in R2 a 30% increase of TVFA, mainly acetate (94%), was observed, thus indicating the simultaneous utilization of hydrogen by the homoacetogens, in accordance with other studies (Kougias et al. 2017). A significant TVFA change in R3 was observed, for both concentration (2.56 gCH3COOH·L−1) and composition: acetate, propionate, iso-butyric and iso-valeric, each accounted for 25% of the total amount, suggesting that a mild inhibition occurred. Moreover, butyrate and valerate isoforms are well known to be specific indicators of process imbalance as reported in the study of Ahring et al. (1995). TVFA increase and accumulation in both reactors determined a slight pH decrease from 7.4 to 7.2. As for R2, specific methane production was almost the theoretically expected (+7%), thus matching the exogenous H2 conversion by hydrogenotrophic methanogenesis. The extra methane expected amount was calculated by applying the stoichiometric conversion of 4 H2 moles to 1 mole of CH4, meaning that all the H2 injected in R2 was consumed by the hydrogenotrophic biomass already present, and only partially by homoacetogens. In fact, Kern et al. (2016), assessed that the hydrogenotrophic methanogenic strain operates below its physiological capacity. By providing different H2 quantities to three different sludge samples, not acclimatized to exogenous H2, they observed, after only 24 h, a positive linear correlation between H2 dose and methane formation rates.

In addition, higher percentages of CH4 in the produced biogas, as well as reduced CO2 contents, were measured in both R2 (73.5% CH4 and 26.5% CO2) and R3 (74.5% CH4 and 25.4% CO2).

Afterwards, during phase III, the H2/CO2 ratio was further raised up to 2:1, corresponding to a dosage of approximately 100 NmLH2·L−1·d−1. As a result, an increase of 13% in R2 and 4% in R3 in the specific methane production, and a decrease in CO2 content were observed. Moreover, the resulting H2 utilization efficiency was evaluated as 98% (Table 3).

Despite the same initial conditions, the two enrichment reactors behaved differently: enrichment dynamic was faster in R2 compared to R3. Moreover, during this phase, VFA composition and concentrations measured in R2 (78% acetate, below 300 mgCH3COOH·L−1) were in the range usually found in well-operating anaerobic reactors (Figure 3). As for reactor R3, TVFA accumulated in the previous phase almost halved (1,090 mgCH3COOH·L−1) and was mainly composed of acetate (60%), thus indicating a progressive adaptation to the new increasing H2 partial pressure conditions. Furthermore a slight pH increase (to about 7.4) in both reactors was observed, confirming that CO2 in the liquid phase was reduced, corresponding to that observed in the headspace gas composition.

During phase IV, an H2 dose of 129 NmLH2·L−1·d−1 was adopted, corresponding to an H2/CO2 ratio of 3:1. A specific methane yield increase of about 30% in both reactors was registered, reflecting a methane content of 77% (Figure 2). Moreover, an effective and stable hydrogenotrophic methanogenic culture enrichment was also confirmed by carbon dioxide content, which was further reduced by 3% in both reactors (22% CO2); also, TVFA concentrations were well below the suggested threshold of 781 mgCH3COOH·L−1 (13 mM of acetate) indicated by Ahring et al. (1995) since acetate was the prevailing component (about 90%).

In the last phase, the stoichiometric value of 4:1 was finally achieved, but no further increments in the specific methane production and in the methane content were observed. However, SHMA values were found to be higher, compared to the previous phases, as will be better explained in the next section. High concentrations of TVFA were again registered in R3, even if this time the accumulation was found to be lower compared to that observed in phase II. Concluding, the acclimation procedure worked better for R2 than for R3, the latter being less stable and efficient. Further investigations are certainly needed to confirm these results and improve the procedure proposed.

The observed differences between reactors R2/R3 and the control reactor R1 during the enrichment procedure (phases II to V) were statistically tested in order to strengthen the main conclusion of the study. The Mann–Whitney U-test showed that there was a significant difference between R1 and R2, as to both the biogas production rate (U = 140, exact significance = 0.003, mean ranks: 9.2 for R1 and 17.8 for R2) and the methane content (U = 160, exact significance = 1.9·10−5, mean ranks: 7.7 for R1 and 19.3 for R2). As for R3, phases II to IV were tested: the methane content measured in R3 resulted in being significantly different from that measured in R1 (U = 96, exact significance = 1.3·10−4, mean ranks: 5.9 for R1 and 15.1 for R3). Conversely, with reference to the biogas production rate, the differences between R1 and R3 were not found to be statistically significant (U = 30, exact significance = 0.243); this is likely due to lower biogas production rate measured in R3 since the beginning of the test. However, it can be stated that the increase of the methane content in biogas due to the enrichment procedure is statistically significant for both R2 and R3 tested against R1.

A COD mass-balance on the three reactors is shown in Figure 4: the balance closed with errors below 10%, allowing validation of the data results of the enrichment trial.

Figure 4

COD mass-balance during the five enrichment phases for the three reactors; R1 is the control reactor; R2 and R3 are the two replicates for the enrichment trial. Percentages indicate the closing errors evaluated for each phase.

Figure 4

COD mass-balance during the five enrichment phases for the three reactors; R1 is the control reactor; R2 and R3 are the two replicates for the enrichment trial. Percentages indicate the closing errors evaluated for each phase.

SHMA tests

SHMA was measured at the beginning of the experiment and at the end of each experimental phase. As reported in Table 4, the incremental hydrogen dosage resulted in an increased SHMA from values of approximatively 80–90 mLCH4·gVS−1·d−1 to values in the range 290–360 mLCH4·gVS−1·d−1. These results are in accordance with Xu et al. (2015) who found, on anaerobic granules from a UASB, an increasing SHMA from 0.2 to 0.6 gCOD·gVSS−1·d−1, corresponding to 70 and 210 mLCH4·gVS−1·d−1. When compared to R2, the lower value measured in R3 at the end of phase II confirms the slight inhibition occurring in that period. Similar conclusions can be drawn by comparing the SHMA to the methane production of the two reactors during phase V when TVFA accumulation in reactor R3 can be observed. In general, results shown in Table 4 clearly indicate that the biomass present in the reactors R2 and R3 was effectively enriched in the content of hydrogenotrophic methanogens.

Table 4

Specific hydrogenotrophic methanogenic activity (SHMA) measured on the effluent digestate at the end of each experimental phase for the three reactors

 SHMA (mLCH4·gVS1·d1)
 
Experimental phase R1 R2 R3 
Start-up 81 ± 36 87 ± 37 91 ± 38 
II H2/CO2 1:1 90 ± 28 105 ± 29 98 ± 27 
III H2/CO2 2:1 85 ± 31 108 ± 34 115 ± 35 
IV H2/CO2 3:1 87 ± 29 234 ± 36 198 ± 29 
H2/CO2 4:1 92 ± 33 359 ± 29 289 ± 38 
 SHMA (mLCH4·gVS1·d1)
 
Experimental phase R1 R2 R3 
Start-up 81 ± 36 87 ± 37 91 ± 38 
II H2/CO2 1:1 90 ± 28 105 ± 29 98 ± 27 
III H2/CO2 2:1 85 ± 31 108 ± 34 115 ± 35 
IV H2/CO2 3:1 87 ± 29 234 ± 36 198 ± 29 
H2/CO2 4:1 92 ± 33 359 ± 29 289 ± 38 

CONCLUSIONS

In this work, the effectiveness of an in situ enrichment procedure was evaluated in terms of process stability, biogas composition and methane production. At the end of the first enrichment phase (H2/CO2 molar ratio 1:1), a significant TVFA accumulation not over 1.8 g·L−1 occurred in one of the two replicates; despite this, the system was able to gain stability, TVFA were consumed and H2 injections were never interrupted. During all the four enrichment phases H2 was converted up to 98%. At the end of the last phase, carried out at the stoichiometric H2/CO2 ratio of 4:1, one of the two reactors was affected by a second stage of slight process instability, causing TVFA accumulation, probably due to the high H2 partial pressure. It is likely that a longer acclimation time is needed when achieving or exceeding the stoichiometric H2/CO2 ratio. However, during the final phase, a methane content percentage of 81% was achieved in one of the two reactors, thus making attractive, in a biological biogas upgrading process operation, the use of an H2/CO2 ratio above the stoichiometric value to boost CO2 conversion, resulting in increasing methane yield and methane content in the biogas produced. Further investigations are needed in adopting or exceeding the H2/CO2 ratio of 4:1 in a longer trial. Furthermore, current results suggest that, up to an H2 dosage below the stoichiometric value, a time duration of 1 week for each increasing step seems to be appropriate for the effective adaptation of the hydrogenotrophic methanogenic culture. A close TVFA dynamics monitoring confirmed this to be a proper tool to follow the running of the anaerobic degradation chain and also anaerobic consortia acclimation to an increasing hydrogen dose. Furthermore, SHMA tests demonstrated the effectiveness of hydrogenotrophic enrichment. Concluding, the enrichment procedure here proposed could be an effective tool for the start-up of pilot- and full-scale reactors to be used for in situ biological biogas upgrading applications.

ACKNOWLEDGEMENTS

The authors wish to thank Dott. Nadia Margariti for her valuable support in the laboratory activities and in the analysis.

REFERENCES

REFERENCES
Agneessens
M. L.
,
Ditlev
L.
,
Ottosen
M.
,
Vinther
N.
,
Lund
J.
,
Jonge
N. D.
,
Holst
C.
,
Vedel
M.
&
Kofoed
W.
2017
In-situ biogas upgrading with pulse H2 additions: the relevance of methanogen adaption and inorganic carbon level
.
Bioresource Technology
233
,
256
263
.
Ahring
B. K.
,
Sandberg
M.
&
Angelidaki
I.
1995
Volatile fatty acids as indicators of process imbalance in anaerobic digestors
.
Applied Microbiology and Biotechnology
43
(
3
),
559
565
.
Angelidaki
I.
,
Alves
M.
,
Bolzonella
D.
,
Borzacconi
L.
,
Campos
J. L.
,
Guwy
A. J.
,
Kalyuzhnyi
S.
,
Jenicek
P.
&
Van Lier
J. B.
2009
Defining the biomethane potential (BMP) of solid organic wastes and energy crops: a proposed protocol for batch assays
.
Water Science and Technology
59
(
5
),
927
934
.
APHA
2005
Standard Methods for the Examination of Water and Wastewater
,
21st edn
.
American Public Health Association/American Water Works Association/Water Environment Federation
,
Washington, DC
,
USA
.
Awe
O. W.
,
Zhao
Y.
,
Nzihou
A.
,
Minh
D. P.
&
Lyczko
N.
2017
A review of biogas utilisation, purification and upgrading technologies
.
Waste and Biomass Valorization
8
(
2
),
267
283
.
Corbellini
V.
,
Kougias
P. G.
,
Bassani
I.
&
Treu
L.
2018
Hybrid biogas upgrading in a two-stage thermophilic reactor
.
Energy Conversion and Management
168
,
1
12
.
Demirel
B.
&
Scherer
P.
2008
The roles of acetotrophic and hydrogenotrophic methanogens during anaerobic conversion of biomass to methane: a review
.
Reviews in Environmental Science and Biotechnology
7
(
2
),
173
190
.
Dou
Z.
,
Dykstra
C. M.
&
Pavlostathis
S. G.
2018
Bioelectrochemically assisted anaerobic digestion system for biogas upgrading and enhanced methane production
.
Science of the Total Environment
633
,
1012
1021
.
ISO (International Organization for Standardization)
1984
ISO 5663:1984 Water Quality: Determination of Kjeldahl Nitrogen—Method after Mineralization with Selenium
.
ISO, Geneva
,
Switzerland
.
Kern
T.
,
Theiss
J.
,
Röske
K.
&
Rother
M.
2016
Assessment of hydrogen metabolism in commercial anaerobic digesters
.
Applied Microbiology and Biotechnology
100
(
10
),
4699
4710
.
Kougias
P. G.
,
Treu
L.
,
Peñailillo
D.
,
Boe
K.
,
Campanaro
S.
&
Angelidaki
I.
2017
Ex-situ biogas upgrading and enhancement in different reactor systems
.
Bioresource Technology
225
,
429
437
.
Lecker
B.
,
Illi
L.
,
Lemmer
A.
&
Oechsner
H.
2017
Biological hydrogen methanation – a review
.
Bioresource Technology
245
,
1220
1228
.
Luo
G.
,
Johansson
S.
,
Boe
K.
,
Xie
L.
,
Zhou
Q.
&
Angelidaki
I.
2012
Simultaneous hydrogen utilization and in situ biogas upgrading in an anaerobic reactor
.
Biotechnology and Bioengineering
109
(
4
),
1088
1094
.
Miltner
M.
,
Makaruk
A.
&
Harasek
M.
2017
Review on available biogas upgrading technologies and innovations towards advanced solutions
.
Journal of Cleaner Production
161
,
1329
1337
.
Muñoz
R.
,
Meier
L.
,
Diaz
I.
&
Jeison
D.
2015
A review on the state-of-the-art of physical/chemical and biological technologies for biogas upgrading
.
Reviews in Environmental Science and Biotechnology
14
(
4
),
727
759
.
Rachbauer
L.
,
Voitl
G.
,
Bochmann
G.
&
Fuchs
W.
2016
Biological biogas upgrading capacity of a hydrogenotrophic community in a trickle-bed reactor
.
Applied Energy
180
,
483
490
.
Ryckebosch
E.
,
Drouillon
M.
&
Vervaeren
H.
2011
Techniques for transformation of biogas to biomethane
.
Biomass and Bioenergy
35
(
5
),
1633
1645
.
Silvestre
G.
,
Fernández
B.
&
Bonmatí
A.
2015
Significance of anaerobic digestion as a source of clean energy in wastewater treatment plants
.
Energy Conversion and Management
101
,
255
262
.
Smith
P. H.
&
Mah
R. A.
1966
Kinetics of acetate metabolism during sludge digestion
.
Applied Microbiology
14
(
3
),
368
371
.
Sun
Q.
,
Li
H.
,
Yan
J.
,
Liu
L.
,
Yu
Z.
&
Yu
X.
2015
Selection of appropriate biogas upgrading technology – a review of biogas cleaning, upgrading and utilization
.
Renewable and Sustainable Energy Reviews
51
,
521
532
.
Ullah Khan
I.
,
Hafiz Dzarfan Othman
M.
,
Hashim
H.
,
Matsuura
T.
,
Ismail
A. F.
,
Rezaei-DashtArzhandi
M.
&
Wan Azelee
I.
2017
Biogas as a renewable energy fuel – a review of biogas upgrading, utilisation and storage
.
Energy Conversion and Management
150
,
277
294
.
Xu
H.
,
Gong
S.
,
Sun
Y.
,
Ma
H.
,
Zheng
M.
&
Wang
K.
2015
High-rate hydrogenotrophic methanogenesis for biogas upgrading: the role of anaerobic granules
.
Environmental Technology
36
(
1–4
),
529
537
.
Zinder
S. H.
1994
Syntrophic acetate oxidation and ‘reversible acetogenesis’
. In:
Acetogenesis
(
Drake
H. L.
, ed.).
Chapman & Hall
,
New York, USA
, pp.
387
415
.