Effect of Hydraulic Retention Time and Organic-Loading Rate on Two-Staged, Semi-Continuous Mesophilic Anaerobic Digestion of Food Waste during Start-Up

Abstract

The start-up of two-stage, semi-continuous mesophilic anaerobic digestion (TSAD) of food waste is stabilized by altering the hydraulic retention time (HRT) and organic-loading rate (OLR). The volumetric biogas yield and composition are studied at OLR (0.25–0.50 gVS/L/d) and HRT (10, 20, 40 days) initiating at OLR 0.25 g VS/L/d and HRT of 20 and 40 days for the respective reactors. Methane (CH₄) from the first stage of the two-staged reactor decreased from 18.20% to 0.06%, fostering hydrogen production in 44 days when the HRT was reduced from 20 to 10 days and OLR increased from 0.25 gVS/L/d to 0.50 gVS/L/d. During the alarming volatile fatty acids (VFA)/alkalinity ratio of 0.76, feeding to the second-stage reactor was halted until pH was restored to 7.00. The restoration of methanogens was evident by an increase in methane from 39.15% to 67.48%. A stable TSAD system produced 22.32 ± 4.16 NmL/gVS and 161.02 ± 17.72 NmL/gVS of yield in respective reactors. Thus, TSAD paves the path for multiple biofuels, i.e., H₂ and CH₄.

1. Introduction

One of the sustainable solutions for the management of the ever-increasing municipal solid waste (MSW) and energy crisis would be energy recovery (ER) from the waste. Among the current approaches for MSW, anaerobic digestion (AD) is an efficient technology because it minimizes greenhouse gas emissions [1] and diverts the organic waste that would otherwise be dumped into landfills to recover energy [2]. In addition, there is increased interest in mature AD processes that enhance biogas production and reduce the use of fossil fuels [3]. Advanced AD processes are yet to be explored through techno-feasibility lenses, and therefore, there have been minimal industrial operations [4]. As biogas can be produced from various types of organic wastes, biological treatments, such as AD, have proven to be attractive options for ER along with waste stabilization. Food waste (FW) is an abundantly available waste biomass that contains high moisture and readily degradable organic matter. Both characteristics make it an attractive feedstock for energy recovery from the AD process [5]. Waste of food is actively connected to the waste of energy and production of greenhouse gases, and various attempts are being made to optimize ER from food waste via AD.

Technically, AD is a matured technology that occurs in the anaerobic condition through a series of processes: hydrolysis, acidogenesis, acetogenesis, and methanogenesis [6]. AD involves psychrophilic, mesophilic, and thermophilic microorganisms which can survive at different temperatures corresponding to 4–25 °C, 20–45 °C, and >50 °C, respectively [7]. However, the conventional single-stage AD process faces challenges, such as system instability, volatile fatty acids (VFAs) accumulation, ammonia inhibition, insufficient buffering capacity, and handling of the high organic-loading rate (OLR) [8]. The two-staged anaerobic digestion (TSAD) concept emerged as a mitigation to these problems, which overcomes such system inefficiencies, thereby improving the biogas yield [9]. The groups of microorganisms involved in the AD process are generally classified into acidogens and methanogens [9]. The first reactor of the two-stage AD produces soluble metabolites, such as organic acids and hydrogen (H₂), and the methane (CH₄) yield in the second-stage reactor can be enhanced by improved fermentation in the first stage [10]. H₂ is generated during the acidogenic and acetogenic stages, while CH₄ is produced in the last stage of methanogenesis by methanogenic archaea. Hence, for these two bacterial communities to function well, a two-staged process is proven to have a higher ER by 21% [11]. Moreover, the two-staged AD process demands critical process control and requires shorter retention times for digestion [12].

Major efforts are being put under the stabilization and development of eco-friendly and sustainable technologies for the upgradation of single and multi-stage methanation systems. One of these efforts has been the co-digestion of FW with added biochar to retrieve FW-biochar-proliferating microorganisms, leading to high AD efficiency [13]. Previous research, on the other hand, suggests that oxygen dosing in a controlled environment can be used to stimulate microorganisms for bioenergy and biochemical production [14]. In addition, multi-stage systems incorporating photofermentation mediated by photosynthetic microorganisms, such as purple non-sulphur bacteria, have improved energy recovery through biological hydrogen production, which has been introduced as an alternative fuel [15]. However, it is advised that phenomena, such as carbon catabolite repression, disrupt the stability of ER [16]. For a stable AD system, various parameters, such as hydraulic retention time (HRT), OLR, reactor volume, pH, and temperature, are considered. Moreover, the dynamics of microbial communities involved, their acclamations to changing operating conditions, and inhibitory conditions in the reactors determine process stability [17,18]. As the adaptation time of the microorganisms and their kinetics rely on the amount and retention time of inoculum added to the substrate, varying HRT and OLR can significantly influence gas yield. A suitable inoculum accelerates the start-up of the reactor system [19] reducing the lag phase of the bacterial community and increasing the digestion rate to enhance biogas production and methane content [20]. The digesters based on mesophilic conditions incorporate mesophiles as the main digesting microorganism and perform ideally around 30–38 °C or at ambient temperatures between 20 °C and 45 °C [7]. Similarly, the enhanced biogas yield is supported by stabilizing operational parameters, such as OLR, HRT, and coupling ratios suitable for the type of substrate used [21]. OLR is one of the crucial parameters that determine the volatile solid (VS) quantity to be fed into the digester each day. Among the parameters determining the microbial community and gas yield, the biochemical pathways are strongly influenced by the operational pH, VFA, and alkalinity [22,23,24]. Similarly, volatile organic acid/total alkalinity (or FOS/TAC ratio) or the ratio of VFA to total alkalinity (TA), helps in monitoring the functionality of a biogas plant as it provides a measure of acidification risk in any reactor [25].

Dark fermentation (DF) is the fermentative conversion of organic substrate to biohydrogen and organic acids, involving a series of biochemical reactions manifested by diverse groups of hydrolyzing and acidogenic microorganisms [26]. DF is led by obligate anaerobes and facultative anaerobes in the absence of oxygen. These anaerobic bacteria act upon the substrate, such as organic waste, and produce hydrogen [26]. DF, sometimes also referred to as acidogenesis, occurs in the absence of light in contrast with the photo-fermentative processes that are driven by photosynthetic bacteria under the influence of light energy [15]. The difference between the DF and acidogenesis processes in AD is that the former is more focussed on biohydrogen production, while the latter is more concerned with the conversion of monomers to VFA which later will be converted to CH₄ in methanogenesis. The products of DF can be utilized in the second stage more efficiently. In addition, pre-treatment of the substrate ameliorates the AD process and increases gaseous yield. By improving substrate biodegradability, increasing soluble substrate amount, and increasing microbial degradation accessibility, pre-treatment has attracted global attention for it drives the metabolic processes towards efficient experimentations of AD and DF [27]. Both biomethane and biohydrogen have positive environmental impacts as efficient energy carriers [4]. Similarly, the soluble by-products typically include acetic acid (17%), butyric acid (28%), propionic acid (5%), caproic acid (11%), and ethanol (1%) [28], which have high industrial value. Additionally, waste biomass is abundant, and hence, DF can serve larger-scale applications [26,29,30].

This research aims to investigate start-up conditions of mesophilic AD of FW for efficient biogas production and to study the effects of various strategies of system stabilization. More specifically, the study evaluates the start-up phase of the TSAD of FW in terms of daily gas production, its composition, VFA production, and alkalinity under the influence of varying availabilities of light, OLR and HRT. The research interprets the results of varying parameters and their responses in terms of the dynamics of the reactors’ performances and how the parameters must be altered to prevent reactor failure. In short, it also marks the necessary steps to be conducted in the reactors for anaerobic digestion to function. With start-up OLRs and HRTs of the experiment of 0.25 gVS/L/d and 20 days in the first stage and 0.25 gVS/L/d and 40 days in the second reactor, respectively, the performance of two-stage AD of FW was studied at OLR of 0.25 gVS/L/d and 0.50 gVS/L/d and HRT (40, 20, 10 days). The reactors were operated for 94 days with various alterations to the operational parameters. The alterations were carried out mainly as stable operational requirements. Characterization of feedstock, inoculum, generated biogas, as well as digestate were carried out throughout the experimental operation. Their performances were monitored daily in terms of biogas production, composition, pH, OLR, HRT, alkalinity, and VFA concentration.

2. Materials and Methods

2.1. Substrate and Inoculum

A food waste (FW) similar to kitchen waste [31] was prepared in the laboratory using 14 ingredients to simulate the energy production from common Asian (Japanese) kitchen waste and to refer to the energy production from the typical composition of waste. The FW was composed of (wet weight basis): 50% vegetables, 20% fruits, 20% rice and noodles, 5% meat, and 2.5% of fish and eggs. The impurities, such as plastic pieces, and neutrals, such as bones and barks, in the substrate sample, were removed before storing at −4 °C for further use. Anaerobic digestate from a functional household-scale AD plant receiving FW was used as inoculum throughout the experimental operation. The inoculation sludge was collected from a 3000 L reactor operating at a mesophilic temperature range.

No pretreatment was adopted for either the feedstock or the inoculum in the first start-up of the experiment. The inoculum, on the other hand, was starved for 36 h before the start of the experiment. The inoculum was stored at a controlled temperature of less than 4 °C. The physical characteristics of the food substrate include dark-bisque color, highly viscous fluid form, and uniform mix of solids. The physical characteristics of inoculum include black color, liquid form, distinct decayed smell, and solids settling with higher density and low viscosity. The physicochemical characteristics of FW and inoculum used in the study are presented in

Table 1. Characteristics of the FW and inoculum used in the study.

Parameters	Units	FW	Inoculum
Total solids (TS)	%	15.94 ± 0.04	0.22 ± 0.01
Volatile solids (VS)	%	14.50 ± 0.16	0.09 ± 0.01
VS/TS	%	90.84 ± 1.20	43.05 ± 0.96
pH		4.36 ± 0.01	8.08 ± 0.01
Total chemical oxygen demand (TCOD)	mg/g	240.4 ± 10.80	NA
Total organic carbon (TOC)	%	8.05 ± 0.09	NA
Total volatile fatty acids (TVFA)	mg/L	NA	364.00 ± 7.00
Total alkalinity (TA)	mg/L	NA	2000.00 ± 16.00
TVFA/TA ratio		NA	0.18 ± 0.01
Density	g/cm3	1.06 ± 0.02	0.99 ± 0.01

NA—not analyzed.

.

2.2. Experimental Setup

The assembly of the two-staged reactors kept in the water bath and the magnetic stirrers are shown in Figure 1 along with their respective daily biogas volume measurement apparatus designed following the study [32]. A 500 mL and another 1000 mL airtight borosilicate bottle (Omsons^®) with graduated polypropylene screw caps and with 400 mL and 800 mL working volume, gas delivery port, and sampling port were used to build the first and the second reactors. The reactors were connected to the buffer apparatus and then to the liquid displacement apparatus for volume measurement. Two magnetic stirrers (Remi^®) with a built-in tachometer were used for continuous stirring for the reactors. The stirrers were operated at 200 RPM. An automatic water bath was set at the mesophilic condition of 35 ± 0.50 °C.

Semi-continuous feeding mode with feeding once a day was adopted for the experiment. Effluent extraction and feeding were carried out maintaining an anaerobic state of the reactors following the process depicted in Figure 1. Fundamentally, the first reactor was designed for facilitating DF for H₂ production and the second for facilitating methanogenesis. During the start-up, the reactors were placed on an opaque layered water bath that only blocked the light to the shoulder level of the reactors. However, the reactors were placed in a dark space, and the neck of the transparent borosilicate glass reactors was left exposed to low light. A predetermined volume of feed and effluent was circulated through the reactors daily with the critical consideration of maintaining the anaerobic state of the reactors. The effluent extraction and feed injection were carried out using an irrigator syringe. The pH, VFA, and TA of the reactors were analysed during the feeding process.

2.3. Analytical Methods

The analytical protocols adopted in the experiments are described. The preliminary characterizations were conducted majorly based on standard methods for the examination of water and wastewater in triplicates [33]. TS and VS content was determined based on the Standard Methods 2540 based on the gravimetric method [33]. Similarly, COD was measured following the closed reflux method [33]. The pH of the samples was sought out using a calibrated pH meter. The pH meter used was HI2020-01 Edge^® by Hanna instruments, which supported 5-standard calibration; pH 4.01, pH 3.00, pH 6.86, pH 7.01, pH 9.18, and pH 10.01. TOC was calculated using the relation, TOC = TVS/1.80, based on the TVS calculated [34]. The pH range of the device was −2.00 to 16.00 with an accuracy of ±0.01 pH at 25 °C. For the determination of VFA, Kapp’s triple point titration was selected as the primary method as it was presented to be the most accurate over a wide concentration range on anaerobic digestate [35]. This method was based on three iterations (pH 4, 4.30, and 5) and measured alkalinity to eliminate errors. TA of the samples was calculated in terms of mgCaCO₃/L by the titrimetric procedure [33]. The density of the FW and inoculum were measured according to the mass present in its unit volume. Daily biogas production was measured by the water displacement method using a 12% NaOH solution to scrub CO₂ from the biogas [36,37]. Biogas composition was confirmed using a portable biogas analyzer (Ruiyi^® Gasboard-3200Plus, China). Standard deviation from the mean is calculated to visualize the likeness and soundness of the experimentally obtained data. The measured volumes of biogas were normalized to standard temperature (0°C) and pressure (1 atm) conditions and expressed as NmL according to the operating temperature of the laboratory when the measurement was done as follows:

$$V_{STP}=\frac{[\left(V_T*273*\left(760-P_W\right)\right]}{\left(270+T\right)*760}$$

where

V_STP = gas volume of standard temperature and pressure (L);

V_T = volume of gas measured at temperature T (L);

T = temperature of the gas or ambient space (°C); and

P_W = vapor pressure of the water as a function of temperature (mm Hg).

3. Results and discussions

With the main objective to analyze the change in the two-stage AD system on varying critical operating parameters, including OLR, HRT, and the recirculation ratio, the laboratory scale reactors were installed and inoculated.

The experiment was initially directed to comparing operational parameters and selecting the most stable parameter. The fundamental strategy for this included the drop in HRT and rise in OLR to seek the optimum of the two operating parameters. However, as the experiments started, a series of responses from the two-stage AD system diverted the experiments’ direction towards rectifying the operational parameters to retain stability, hence, making the parametric changes a requirement of the system itself rather than being imposed for a comparative study. This series of parametric change requirements of the system enabled analysis of the effects of the changes and also indicated that the two-staged AD system is spontaneous rather than planned. However, the causes for the system to require a series of changes are identified clearly from the daily analyses conducted in the system. The major events and operational variations, including feed halt, change in HRT, repositioning of reactors to avoid the light and change in OLR, are presented according to the reactors’ requirements along with the timeline in chronological order.

The performance of two-stage AD of FW was studied at OLR of 0.25 gVS/L/d and 0.50 gVS/L/d and HRT (40, 20, 10 days). In the first reactor, 20 days of HRT equalled a circulation rate (Q) of 20.00 mL/d in the 400 mL working volume. This circulation rate consequently fixed the minimum HRT of the second reactor to 40 days due to its 800 mL working volume. According to an OLR of 0.25 gVS/L/d, 690.71 mg of FS was diluted to 20.00 mL and fed into the reactor. The reactors were operated for 94 days, and their performances were monitored in terms of biogas production, composition, pH, OLR, HRT, alkalinity, and VFA concentration. The average pH of the feedstock fed to the first reactor over the course was 4.28 ± 0.16.

Figure 2a,b represent the daily gas production from the first and second reactors, respectively. As seen in Figure 2a,b, there was no circulation of feed and effluent in the first week. This is done to facilitate the growth of microorganisms in the reactors [38]. Figure 2a,b also show that the yields are not significantly stable according to the change in feeding, OLR, and HRT throughout the operation. This is because the changes were mostly imposed as the requirement of stability of reactors rather than as comparative experiments. The consequences that required the parametric changes are listed along with the responses in

Table 2. Summary of the major parametric changes, the reason behind making the change, and the response of the TSAD system towards the change.

Parametric Changes	Cause (C) and Response (R) of Change
First-stage reactor
OLR increase to 0.50 gVS/L/d	R: Methane % increased (18%, 26th day)
HRT reduced to 10 days	C: To wash away methanogens R: Methane % decreased (0.90%, 48th day)
Reposition the reactor to an adequately lit space	R: Inhibition of photo-fermentative bacteria R: pH decreased (3.70, 49th day)
Addition of an opaque layer of wrap	C: To halt the growth of photo-fermentative bacteria R: pH increased (4.80, 94th day) R: Unstable yield in this period
Second-stage reactor
OLR increase to 0.50 gVS/L/d	R: Low increase in methane % (18.00%, 26th day) R: pH decreased (6.90, on the 28th day from 7.70, on the 18th day)
OLR decreased to 0.25 gVS/L/d (28th day); HRT unaltered.	C: HRT of the first reactor decreased to 10 days The coupling ratio increased to 4 R: pH decreased (6.50, day 33)
Feeding paused (34th–40th days)	C: Alarming VFA:TA ratio (0.77, 33 days) R: pH restored to 7.20, reactor stabilized
Reposition the reactor to an adequately lit space	R: Invasion of pink photo-fermentative bacteria
Addition of an opaque layer of wrap	C: To limit the growth of photo-fermentative bacteria R: Microbial counteraction for dominance R: Unstable yield data
A restart of the TSAD system
Inoculum pre-treatment in the first-stage reactor (90 °C, 45 min)	C: To deactivate methanogens R: Absence of methane% in starting phase
The second reactor volume upgraded to 1.80 L	C: To permit maximum ER R: Second reactor stable with entire effluent from 0.4 L first reactor
Recirculation started (on the 13th day) Recirculation lowered (50.00%, 25th day)	C: Alarming pH in the first reactor (3.20, 12th day) R: Attained stability (30th–40th day)

. These parametric changes cause the changes and respond to the changes that are comprehensively enlisted in Table 2. Figure 3 represents the change in pH and VFA to alkalinity ratio in both reactors.

The composition of the biogas in Figure 4a obtained from the first reactor within 90 days predicts the possibility of H₂ gas from the first reactor, considering the low percentage of CH₄ and CO₂ obtained [39,40].

In Figure 4a, 25–30 days demonstrate there is CH₄ production in about 17% of the biogas produced implying that methanogenic bacteria are present in the first reactor. However, the presence of CH₄ in the first reactor is seen to decrease to a null value over time, especially after 55 days of operation. On the other hand, Figure 4b shows the increase in CH₄ composition in the second reactor over time. It is visible that the CO₂ content of biogas produced from the second reactor is less than the first reactor.

As seen in Figure 2a,b, the reactors started producing biogas right after they were set up. Around 88 mL/gVS and 163 mL/gVS of biogas are produced from the first and second reactor, respectively, without any feeding or circulation during the first four days of operation. This signifies that the inoculum was active, and the lag phase of bacteria was short. This is valid because the inoculum was collected from an operational AD plant. The starting OLR and HRTs of the experiment were set to 0.025 gVS/L/d and 20 days in the first stage and 40 days in the second reactor, respectively, to facilitate the housing and growth of essential microbial life in the reactors. During the period of stabilization, a very small FW OLR of 0.025 gVS/L/d was added for acclimatizing, housing, and facilitating essential microbial growth in the reactors. It is observed from Figure 4a that a small fraction (7.21%) of CH₄ was present in the composition of the biogas produced from the first reactor. This is because the inoculum used was derived from an AD plant operational for methanogenesis. Then, the biogas composition from the second reactor comprised only 5.0% CH₄.

The OLR increased to 0.25 gVS/L/d, which indicates the initiation of feeding from the eighth day of initiation. The HRT of the first reactor was set to 20 days and that of the second was set to 40 days. After 10 days of evaluations of these parameters, it was seen that the daily specific biogas yield while feeding an OLR of 0.25 gVS/L/d averaged 423.60 and 205.70 mL/gVS/d.

The FOS TAC ratio when seen at 0.07 indicates that more of the substrate can be fed into the digester because the threshold of stability is not seen (Wang et al., 1999). This study had a ratio of 0.42 to 1.00 and maintained stability with maximum production at 0.980. The OLR of the reactors were then increased to 0.50 gVS/L/d from the 18th day of installation as seen in Figure 2a,b. A similar study [32] suggests the total alkalinity in the system should be higher than the total concentration of volatile acid.

The change in pH during these phases of reactors is also significant and is recorded in Table 2 as well. From Figure 3 it can be observed that the pH values of both reactors were around 7.50 during the initiation and dropped down over time. The pH records of the first and second reactors on the 18th day before the increment of OLR were 5.30 and 7.68, respectively. This record of pH is considered beneficial for targeted bacterial growth. The decrease in pH of the first reactor signifies that it is in favor of dark fermentative bacteria. The lower pH favors the production of H₂ as suggested by [26]. The stable pH of 7.50 in the second reactor also facilitates the growth of methanogens as it is habitually seen that methanogenic bacteria require a pH value ranging from 7.00 to 8.50 [32,36]. Chronologically, the performances of both TSAD reactors are evaluated individually in the following sections.

3.1. Performance Evaluation of the First Stage after 18 Days of Installation

After an increment of OLR from 0.25 gVS/L/d to 0.50 gVS/L/d on the 10th day of operation, there is no significant increase in biogas yield. This might be due to the influence of the methanogenic microbial community in the first reactor. From Figure 4a, the biogas composition analysis conducted on the production from the first reactor on the 20th to 26th days shows the rise in CH₄ composition from around 5% to 18%. Peculiarly, methanogens survived on a very low pH of 4.60, as shown in Figure 3. The drop in pH accompanied by methanogenic bacterial growth indicated a contrast and disturbance in microbial communities present in the first reactor. Short HRT usually results in the accumulation of VFAs, whereas at longer HRT, digester components are not effectively utilized [40]. This might also be the reason for the diminished biogas production. Recirculation of effluent from the second reactor could be the solution for both reactors in cases such as these [41]. However, as the first reactor showed an unusual growth of methanogens, further addition of methanogenic effluent from the second reactor would result in methanogens dominating the first reactor, which was fundamentally designed to be dark fermentative. Thus, wiping the methanogens out of the first reactor received the highest priority and urgency as well. As mentioned in Table 2, the HRT of the first reactor decreased to 10 days, increasing the circulation volume from 20 mL to 40 mL. The HRT of the second reactor also decreased to 20 days from 40 days.

The decrease in HRT resulted in the successful washing out of the methanogenic bacterial community from the first reactor. The CH₄ fraction of gas composition during the period of HRT change was 17.72%. From Figure 4b, it is seen that CH₄ composition increased and decreased to 0.90% on the 48th day and remained around that level throughout the operational period until the 94th day when the composition of CH₄ was only 0.02%. This signifies the successful wash-out of methanogens from the first reactor that was constructed fundamentally for DF.

Although the methanogens were washed away from the first reactor by the 48th day of operation, stable values of yield could not be obtained until the 94th day with the same operating parameter of 10 days HRT and OLR of 0.50 gVS/L/d. This might be a consequence of the repositioning of the reactors to a bright laboratory on the 29th day, as indicated in Figure 2a. However, the reactor was wrapped with an opaque layer on the 48th day of operation. This phase of 19 days indicates the invasion of photo fermentative bacteria in the reactor. It can also be seen in the pH record that the pH dropped down to 3.70 from 4.20 in this phase. This is seen followed by a fluctuation of pH with the lowest value being 3.60 after the wrapping of the opaque layer. The period of operation after that shows a variation in pH increasing to 4.80 on the 94th day of operation. This inconsistency of pH indicates the instability of the reactor and thus, explains fluctuating yield results from the first reactor. This inconsistent period might be the effect of the contrasting nature [26] of the inhabited bacteria communities that naturally struggle for dominance [30].

3.2. Performance Evaluation of the Second Stage after 18 Days of Installation

In the second reactor, the CH₄ composition was seen to increase from 4 to 18% during days 20–26. This is a low fraction of CH₄ composition considering the pH condition of around 7.80 in the reactor. Although the neutral pH suits the growth of methanogenic bacteria, the biogas composition does not indicate the dominance of methanogen growth. This might be due to the growth of photo-fermentative microorganisms, as mentioned in Table 2, and later witnessed in the experiment duration, as presented in Figure 5.

Upon increasing the OLR to 0.50 gVS/L/d, the pH seemed to drop rapidly from 7.68 to 7.00 in the period of the 9th to 19th day, as seen in Figure 3. This pH drop corresponded to a rise in VFA from 200.00 to 622.00 mg HAc/L which resulted in an increase in the VFA/alkalinity ratio from 0.07 to 0.40. This is a considerable increase in the ratio compared to the stable range of 0.10 to 0.40 [42]. The total alkalinity, however, was constant with values around 1400 mgCaCO₃/L which lay in the optimum working range for methanogenic AD [30].

As the first reactor required an HRT drop for washing out methanogens, the volumetric feeding rate increased for the second reactor as well, decreasing its HRT to 20 days. Although the VFAs were increasing rapidly with the increase in OLR, their values lay within the optimum range. Hence, the coupling ratio of the reactors was not disturbed, and HRT decreased proportionately. However, this resulted in an alarming condition for the second reactor. The pH is seen to decrease from 7.00 to 6.50, which is presented in Figure 3. The VFA/alkalinity ratio elevated to 0.70 from 0.30 as a result of the HRT decrease. This value indicates instability as it is far beyond the optimum working range of 0.07 to 0.40, posing an acidification threat in the reactor system [42,43]. Therefore, feeding was stopped to the second reactor for eight days starting from the 34th day of installation. This resulted in a decrease in VFA and an increase in pH from 6.50 to 7.20. Moreover, the VFA/alkalinity ratio is seen to have restored to the optimal range after the stoppage in feed. This indicates that stopping the feed could restore the stability of the AD system with a higher range of VFA accumulation.

There was a cumulative gas production of around 470 mL in the period of absence of feed from the second reactor, which indicates that there was undigested FS in the reactor volume (Figure 2b). This adds up to the analysis that the coupling ratio of the first-stage working volume to that of the second-stage in a coupled system of DF and methanogenic AD should be more than 1:2. It is observed that the total effluent of the first fermentation reactor was excessive as feed to the second reactor, which could result in reactor failure [44,45].

The HRT of the second reactor was increased to 40 days, and OLR was decreased by diluting the effluent of the first reactor after the 41st day of the operation. Although the reactors’ characteristic readings, such as pH, VFA, and alkalinity, were seen to be in the optimum range after the 41st day, the specific yield from the reactor was not stable. This is suspected as the result of the invasion and growth of the pinkish-red pigmented photo-fermentative bacterial community in the reactor. The non-native bacterial growth was favored by the repositioning of the reactors to a bright laboratory on the 29th day, as indicated in Table 2 and seen in the timeline in Figure 2a. As the new position had adequate daylight, a change in reactor color was noticed in the second reactor. A figure depicting the color change in the second reactor, as seen in Figure 5, indicates the invasion of the photo-fermentative microbial community. The color change was significantly noticed from straw white to pink and finally to maroon, as seen in Figure 5. The collage of photo records labelled 1 to 13 in Figure 5 represents the accurate color and pigmentation alterations of the effluent from the second reactor from the 29th to the 55th day after installation, respectively.

As soon as the change in reactor color was noticed, the reactor was wrapped with an opaque layer on the 48th day of operation [38]. Based upon the performance of the reactor and the visibility of dominant populations in the reactor, the reactor had been characterized by the invasion of photo-fermentative bacteria [26], more so as the consortium could alter the dynamics of the reactor system [46]. The unstable yield had been followed by its presence in response to the shift in reactor position as compared to the influence of bacterial growth in biogas yield [47]. The photo-fermentative bacteria seemed to compete with the methanogens for the VFA. The fluctuation and decrease in yields shown in Figure 2a,b justify the competition for the substrate posed by the photo-fermentative bacteria. The instability caused by the bacteria made it even more difficult to wipe them off the methanogenic second reactor because decreasing the HRT to increase the flow rate would risk the methanogens being wiped off before the photo-fermentative bacteria community [46,47]. This summarizes the evaluation of the TSAD reactor system 94 days from installation.

3.3. Performance Evaluation after System Restart

It was seen that a coupling ratio lower than 1:2 (i.e.,larger working volume of the second stage) is to be adopted for optimum ER from the selected substrate and system. Hence, a similar new system was restarted with working volumes of 0.40 L and 1.80 L in the first and second reactors sized 0.50 L and 2.0 L, respectively, for further investigation. The new system was operated for 40 days to record a stable response in gas yield, FOS TAC ratio, and pH. This operation was carried out without the provision of light entering the reactors. The response is analyzed based on Figure 6 and Figure 7. Figure 6 presents the record of daily biogas yields from both the new reactors along with the change in HRT and OLR carried out through the operation. The change in the FOS TAC ratio of the second reactor and the pH of both reactors is presented in Figure 7. The major operational modifications, as seen in Figure 6 and Figure 7, include two changes each in HRT (10 days from 20 days in the first reactor after the 18th day, corresponding to 45 days from 90 days in the second reactor) and OLR (0.500 from 0.250 gVS/L/d after the 19th day) and a change in the recirculation amount after the 11th day from restarting the system.

The new system of a two-staged DF–AD couple was started with the same feedstock and inoculum for the second reactor. However, for the first stage, the inoculum was pre-treated with heat at 90 °C for 45 min to inactivate any methanogenic bacteria and isolate the spore-forming fermentative bacteria suitable for DF [48]. As the reactors were wrapped by an opaque layer from the beginning, the instability due to the invasion of any photo-fermentative bacterial community was not visible throughout the second operation. However, the early 10 days of operation showed an unstable response, as seen in Figure 6, possibly due to the microbial community lacking acclimatization. The first five days of operation produced significant amounts of biogas (around 75 mL/gVS addition) from the first stage, as seen in Figure 6, which was followed by a halt in gas production for the next five days. This halt in gas production was accompanied by a fluctuation in gas yield and a rise in VFA’s concentration in the second reactor resulting in a relatively higher FOS TAC ratio, as seen in Figure 7, as well as the first reactor pH drop to 3.20 on the 12th day from the restart.

As a natural strategy for stabilizing the operation, the second reactor’s effluent was recirculated after the 13th day. Pausing the feed circulation or adding bases to the first reactor to stabilize pH were the secondary strategies for their strident nature. As recirculation enables the injection of methanogens into the first reactor, the risk of instability was still prevalent. However, as the pH was only 3.57, it seemed unfavorable for the methanogens’ growth and dominance [32,49]. This was confirmed by the gas analysis conducted in that period which recorded a 0% CH₄, 8% CO_2, and 852 ppm H₂S concentration in the gas produced from the first reactor. Meanwhile, the second reactor showed a 12.95% CH₄, 7.58% CO₂, and 31.13 ppm H₂S concentration. After the OLR was increased to 0.50 gVS/L/d on the 18th day, an increase in CH₄ composition in biogas from the second reactor was recorded. The FOS TAC ratios as well as the pH of both reactors were stable after the 25th day of a system restart. As stable responses were recorded from the reactors, the recirculation was decreased to 50%. Though the highest productions of H₂ and CH₄ are not noticed under recirculation, as seen in a similar study [50], it is noticed that recirculation helped in gaining stability, and reducing recirculation did not alter the stability of the reactors, as suggested by [50,51]. This fact is justified by the record of the first reactor retaining its pH and stability even after recirculation decreased. The period from the 30th to the 40th day after restart of the system exhibits stable pH, FOS TAC ratio gas yield, and composition as well. In this period, at an OLR of 0.50 gVS/L/d and HRT of 10 and 45 days, an average of 22.32 ± 4.16 NmL/gVS and 161.02 ± 17.72 NmL/gVS of biogas yield was observed in the first and second reactors, respectively.

3.4. Considerations of this Study on FW-Based, two-Staged AD System Startup in the Context of a Circular Economy

AD can be optimized when the feedstock is based on organic waste due to the reduction of feedstock cost. This encourages AD practices in the context of developing countries, such as Nepal. One of the most readily available and efficient feedstocks for biogas production is FW. The stabilizing strategies and variation of operational parameters sought in this study could aid in the further development of standard parameters for scale-up FW digesters. The outcomes of this study can aid as fundamentals to further research in single-staged, two-staged, as well as multi-staged AD of FW, which could include the effects of various pre-treatment methods, physical operational alterations (temperature, stirring mechanism, etc.), co-digestion, or even coupling with electrical upgradation systems [52,53]. The rate of stable biogas production from FW obtained in this study can be used to compare with large-scale plants and their efficiency.

The addition of chemically synthesized externalities in an AD reactor could result in harsh instability in microbial communities involved leading to serious complications, such as microbiology washout, encrustation, etc. [54,55]. Hence, studies as such can help detect the operational fault and revert to solvable steps in biogas plants using feed recirculation strategies.

Biogas plants treating the organic waste can mitigate the climate change effect as they reduce the production of gases, such as CH₄ which are emitted into the atmosphere from landfills and instead use as fuel. Manifested biogas is a vital part of the energy supply system and waste management chain, with the ability to fulfil all energy needs of modern society via upgradation. The leftover slurry (digestate) from the AD plant improves soil quality for agriculture, enhances food production, and can be used commercially used as a bio-fertilizer. In addition, resourceful biochemicals like propionic acid, butyric acid, and caproic acid are secondary by-products of the first stage of TSAD.

Obtained energy can be directly used with combustion engines, fuel cells, or even converted into electrical energy.

A fraction (10–25%) of H₂ when mixed with (70–85%) CH₄ forms hythane. Trademarked by Eden Energy, hythane has the advantages of low carbon emission and high-performance efficiency. The combination of inflammable hydrogen and slow-burning, hard-igniting methane is considered ideal [56,57]. Bio-hythane produced via a two-stage process is greener than hythane produced using natural gas. As multi-stage AD holds the ability to produce both gases in high purity and fitting proportion, there is a possibility of the positive use of this system for energy recovery from waste.

4. Conclusions

The operation of a TSAD system consisting of the couple of DF in the first stage and methanogenic AD in the second is stabilized by altering natural parameters, such as the recirculation ratio, HRT, and OLR. The coupling ratio higher than 1:2 is seen to be favorable for complete ER from the FW. The increase in flow rate in the first reactor by decreasing HRT assisted in washing out methanogens in the first reactor. The provision of light in the reactors might lead to the growth of photo-fermentative bacteria that compete with methanogens VFA consumption, thus reducing the biogas yield and increasing instability. Recirculation of effluent is witnessed as an effective natural approach to improving the performance and stability of the TSAD system. During stable operation of the TSAD system at an OLR of 0.50 gVS/L/d and HRT of 10 and 45 days, an average of 22.32 ± 4.16 NmL/gVS and 161.02 ± 17.72 NmL/gVS of gaseous yield was observed in the first and second reactors, respectively. pH tests and VFA: alkalinity tests of the reactor sample are observed to be significant ways to analyze the stability of the TSAD process. The feeding alterations conducted as per the needs of these tests can assist in maintaining a healthy reactor system. The reactor system can be further evaluated and improved with analysis of soluble metabolites so that effective monitoring of intermediate products in the two-stage system is made. In addition, hydrogen gas analysis will determine the composition of gas produced from the one-stage reactor. Incorporating change in the bacterial community also would help monitor the metabolic pathways and performance of the two-stage process.