Food Waste Bioconversion Features Depending on the Regime of Anaerobic Digestion

Abstract

Approximately one-third of global food production is wasted annually, which contributes significantly to greenhouse gas emissions and economic costs. Anaerobic digestion (AD) is an effective method for converting food waste into biogas, but its efficiency depends on factors such as temperature and substrate composition. This study compared mesophilic and thermophilic AD of selectively collected fruit and vegetable waste, quantifying process efficiency and identifying factors leading to collapse. Studies were performed in 1 dm³ reactors with gradually increasing organic loading rates until process collapse. Process dynamics, stability, and gas yields were assessed through daily biogas measurements and analyses of pH, FOS/TAC ratio, sCOD, ammonia, volatile fatty acids, alcohols, total and volatile solids, and C/N ratio. Research has shown that peak methane yields occurred at OLR = 0.5–1.0 kg VS·m⁻³·d⁻¹, with thermophilic systems producing 0.63–5.48% more methane during stable phases. Collapse occurred at OLR = 3.0 in thermophilic and 4.0 in mesophilic reactors, accompanied by sharp increases in methanol, acetic acid, butyric acid, propionic acid, and FOS/TAC. The pH dropped to 5.49 and 6.09. While thermophilic conditions offered higher methane yields, they were more susceptible to rapid process destabilization due to intermediate metabolite accumulation.

1. Introduction

Approximately one-third of global food production is wasted each year, amounting to about 1.354 billion tons. This loss costs the global economy more than $940 billion annually. The United Nations Sustainable Development Goal (SDG) Target 12.3 aims to reduce food loss and waste by 50% by 2030. Food production contributes significantly to carbon emissions at each stage of the supply chain: growing, harvesting, storing, processing, transporting, cooking, and discarding. An estimated carbon footprint of food production is approximately 3.3 billion metric tons of CO₂ equivalent, contributing 11.8% of global greenhouse gas emissions [1,2,3].

According to FAO, 14% of food globally is wasted after having been harvested and before going on sale, and the next 17% of waste occurs in retail, in food service, and by consumers [1]. Data collected by Eurostat show that an average Polish individual wastes 106 kg of food annually, while the European average is 131 kg per person [4]. Assuming all these amounts can be subject to selective collection, it is, in total, 4 million tons of biomass per year. Hence, raw materials from the food industry in Poland are growing in importance. In 2024, they contributed to as much as 69% of substrates used to produce agricultural biogas [5].

The most frequently wasted plant-based products are fruit and vegetables (45%) and grains (30%) [1]; however, almost all organic matter formed while producing, distributing, and consuming food can be reprocessed for the food industry and energy generation purposes [6,7,8]. Considering the latter, key factors that decide the affordability of using the particular substrate are its price, biogas efficiency, and the cost of transport to the biogas plant.

The generation of food waste and its further utilization are problematic for many reprocessing plants [9,10,11]. Such a type of waste is easily biodegradable, both in aerobic and anaerobic conditions. It is crucial since its energetic value decreases, while its potential environmental hazard increases over time [12]. When being uncontrolledly stored, the emission of gases polluting the atmosphere occurs [12], which in extreme cases can lead to fatal poisoning [13]. Waste that is subject to prismatic storage releases leachate, which possibly results in the pollution of surface water and groundwater [14]. Additionally, expired food that is the product of food spoilage is likely to undergo a fierce foaming process, thus it needs to be stored in containers in which intense stirring is provided.

On the other hand, the susceptibility of food waste to biodegradation is positive in terms of biological methods of waste treatment due to the possibility of using it efficiently both in composting as well as methane fermentation [15,16].

Anaerobic digestion (AD) as a bioconversion method is well-known and widespread. Different types of wastes have already been studied, ranging from wastewater [17,18,19,20] or sewage sludge [21,22,23,24], through animal manure [25,26,27] and agricultural waste [28,29,30]. Various studies also included the topic of co-digestion [31,32,33,34,35,36]. Numerous pre-treatment methods used to intensify the fermentation process were reviewed and discussed by Sinan Akturk and Goksel [37] and others [38,39,40].

Some studies were also conducted in terms of fruit and vegetable waste. For instance, research concerning kinetic parameters of AD of organic waste from the fruit and vegetable processing industry in a continuous stirred tank was conducted by Streitwieser [41]. It showed that the rate of anaerobic fermentation varies, depending on the organic load and temperature, and the rate of degradation is higher in the thermophilic regime. The quantity, composition, and methane potential of different food wastes were summarized in [42]. Chavan et al. [43] investigated the thermochemical pretreatment of kitchen waste and showed that it allows for achieving high levels of glucose in the mixture of kitchen waste, facilitating the process of enzymatic hydrolysis. Some studies on AD of food waste with the addition of micronutrients have shown that the addition of micronutrients reduces the accumulation of long-chain fatty acids, stabilizes the fermentation process, and increases the methane yield by 15–65% [44]. Mono-digestion of food waste often leads to digester instability and even collapse at higher organic loading rates (OLRs), above 2.5 g of volatile solids (VS)∙dm⁻³∙d⁻¹), especially under thermophilic conditions, due to the accumulation of volatile fatty acids and ammonia inhibition [45,46,47]. However, a clear gap remains—no previous studies have quantified metabolite build-up preceding process collapse at two different temperature regimes, despite temperature being one of the most critical parameters influencing anaerobic digestion performance. The temperature regime of AD has a significant impact on the process and has been studied in detail by different authors [21,48]. In industrial conditions, biogas plants work mainly in mesophilic systems (the vast majority of European biogas plants) but also in thermophilic systems; for example, biogas plants in Denmark and most municipal waste treatment plants. Both options vary significantly, for example, in average biogas and methane production yield, organic matter decomposition rate, and the process’s stability and sensitivity to the presence of inhibitors [49,50,51,52]. According to [53], thermophilic fermentation of food waste is characterized by higher levels of biogas productivity, yield of fatty acids, pH, and concentrations of chemical compounds, as well as greater efficiency of organic compound removal. Due to the high diversity of fermentation microflora [54], it is not possible to identify one optimal technology for all substrates. An increase in the process temperature can result in higher biogas efficiency of straw substrates and shorter attenuation time [55,56]. On the other hand, it also influences increasing susceptibility to high concentrations of long-chain fatty acids [57,58]. In terms of easily decomposable and finely fragmented substrates, there is no economically affordable way to conduct the process in thermophilic conditions [54,59].

The composition of food waste also influences its bioconversion features. The typical composition of food waste varies a lot depending on different characteristics (source, region, dietary habits, etc.). Generally, it includes a mixture of carbohydrates in high concentrations, proteins, fats, fiber, and water [60]. Typical components of food waste are summarized in

Table 1. Typical components of food waste (authors’ meta-analysis).

Component	Volume	Sources	Types
Carbohydrates	15–30%	Fruit, vegetables, cereals, bread, pasta	Sugars, starches, cellulose
Proteins	5–10%	Meat, dairy products, legumes	Amino acids, peptides
Fats	5–15%	Oils, butter, dairy products, meat	Fatty acids, glycerol
Fiber	5–20%	Vegetables, fruits, whole grains	Lignin, cellulose, hemicellulose
Water	60–80%	All types of food waste	Microelements and other ingredients

. Additionally, food waste also contains minerals (calcium, magnesium, potassium, sodium) and vitamins.

The suitability of food waste for anaerobic digestion is determined by several key factors. A high content of carbohydrates and proteins, which are easily biodegradable by methanogenic microorganisms, promotes efficient biogas production. The presence of fats, due to their high energy value, can significantly enhance biogas yield; however, excessive amounts of fats may inhibit the anaerobic digestion process. A high water content facilitates the transport and mixing of substrates within the digestion reactor, meaning that food waste typically does not require additional dilution. Furthermore, the presence of trace elements such as calcium, magnesium, and potassium is essential for sustaining the lifecycle of microorganisms in the reactor.

Food waste typically contains small amounts of toxic substances and a balanced carbon-to-nitrogen ratio (C/N) that is beneficial for the stability of the AD process.

Due to the above, the composition of food waste makes it a suitable substrate for anaerobic digestion, promoting efficient biogas production and the biodegradation of organic matter.

The aim of this study is to compare the efficiency of anaerobic digestion of food waste in mesophilic and thermophilic conditions and to determine possible reasons for the collapse of the process. To achieve this goal, the course of the study included determining the dynamics of the AD process and the efficiency of the biogas per kg of VS (volatile solids) of substrate, as well as identifying the limitations of the AD process.

2. Methodology

2.1. Experimental Setup and Analytical Methods

The experiment was conducted through anaerobic digestion in a set of multi-chamber biofermentors constructed in the Laboratory of Ecotechnologies and described by Cieślik et al. and Waliszewska et al. [55,61]. Anaerobic digestion was carried out in six reactors (three replications) with a working capacity of 1 dm³.

The mesophilic fermentative inoculum was obtained by separating the liquid fraction of the digestate pulp from an operating agricultural biogas plant. Thermophilic inoculum was obtained by rapidly increasing the temperature from 39 °C to 52 °C and leaving it for a few weeks to multiply the thermophilic microorganisms.

Reactors were fed once a day. Initial OLR was 0.5 kg VS∙m⁻³∙d⁻¹ and it was gradually increased by 0.5 kg VS∙m⁻³∙d⁻¹ after the gas production stabilized until the process collapsed. Qualitative and quantitative measurements of produced biogas were carried out before the beginning of the reactor feeding procedure.

Process dynamic and stability were determined by the following parameters: pH of digested pulp, FOS/TAC ratio (FOS—volatile organic acid concentration; TAC—total alkaline carbonates), concentration of sCOD (soluble chemical oxygen demand), ammonia, VFA (volatile fatty acids; lactate, formate, acetate, propionate and butyrate), alcohols (ethanol and methanol), TS (total solids), and VS of the digested pulp.

The above-mentioned measurements were conducted as follows: pH was determined according to PN-90 C-04540/01; FOS/TAC ratio was measured by titration using an automatic TitroLine^® 5000 (SI Analytics GmbH, Mainz, Germany) with 0.1 N H₂SO₄ [62]; total solids (TS) were analyzed in accordance with PN-75 C-04616/01; and volatile solids (VS) were determined according to PN-Z-15011-3.

For sCOD, ammonia, VFA, and alcohols, post-fermentation samples from three biological replicates were pooled, centrifuged at 15,000 rpm for 10 min, diluted 10–100×, and filtered through a 0.45 µm hydrophilic PTFE syringe filter. All analyses were performed on the resulting composite sample. sCOD was determined using the dichromate method with Lovibond^® ready-to-use kits. Ammonia was analyzed using the VARIO Powder Pack (Lovibond^®, salicylate–cyanurate method, 10 mL cuvette test produced by Tintometer GmbH, Dortmund, Germany). VFAs and alcohols were quantified by high-performance liquid chromatography (HPLC, Agilent Technologies 1200 series, Santa Clara, CA, USA) equipped with a refractive-index detector and a Rezex ROA-Organic Acid H⁺ column (Phenomenex Inc., Torrance, CA, USA), using 0.001 M H₂SO₄ as the mobile phase under isocratic elution at 0.6 cm³ min⁻¹ and a column temperature of 40 °C.

Parameters such as daily volume, composition of the biogas, and the yield of methane and biogas per Mg of VS of substrate were determined. The C/N ratio of food waste was determined using an Organic Elemental Analyzer Flash 2000 (Thermo Fisher Scientific, Cambridge, UK). The methodology of determining the above parameters has been described previously by Cieślik and Kozłowski [55,63].

All charts were prepared using Microsoft Office tools (Microsoft Corporation, Redmond, WA, USA).

2.2. Feedstock Used

The food waste used in this study was obtained through selective waste collection, meaning the manual separation of vegetable and fruit residues from other waste streams at households and local markets in the Greater Poland region. The collected feedstock was delivered as a mixed batch from a communal container dedicated exclusively to vegetable and fruit waste. As the mixture was already combined at the collection point, it was not possible to determine the exact proportion of individual fruit and vegetable components. The substrate was subsequently shredded using an industrial food shredder (Wilk) and homogenized in a blender to form a slurry. The prepared feedstock was then divided into smaller portions and stored at −20 °C until use. The total solids (TS) content was 15.70%, with volatile solids (VS) accounting for 90.7% of TS. The carbon content was 1.88% TS and the nitrogen content was 38.94% TS, resulting in a C/N ratio of 20.66.

3. Results

3.1. Parameters of the Substrate

The food waste used in the experiment was characterized by a TS value of 15.70%, which allowed for conducting the continuous methane fermentation process without the necessity of substrate dilution. Too high a percentage of dry matter could result in compacting fermentation pulp and, in turn, difficulties in stirring the contents of the reactor and loss of pumpability of the fermentation pulp. Moreover, when percentages of TS and VS are 15.70 and 90.70, respectively, even with a relatively high OLR of 5 kg VS∙m⁻³∙d⁻¹, HRT (hydraulic retention time) would be 28 days.

The next reason for using food waste consisting of fruit and vegetables as a substrate in the process of methane fermentation is its C/N ratio, which is 20.66.

3.2. The Dynamics of the Biogas Production

Continuous methane fermentation of food waste was carried out for 154 days in mesophilic conditions and 104 days in thermophilic conditions. The duration of each variant was defined by the onset of process instability, indicated by reactor acidification and complete cessation of gas production. Figure 1 shows the daily production of methane and biogas over time, depending on the OLR type. Figure 2, Figure 3, Figure 4, Figure 5 and Figure 6 present changes in process parameters (pH, FOS, TAC, FOS/TAC, sCOD, the concentration of metabolic intermediates, and production efficiency) depending on OLR.

While conducting the process in different conditions, significant changes were observed. Initially, the gradual increase in the daily production of methane and biogas was noted with increasing OLR at both 39 °C and 52 °C. In both types of conditions, initial OLR at the level of 0.5 kg VS·m⁻³·d⁻¹ was maintained for 14 days. After that time, the stability of biogas and methane production was achieved, and the OLR started to be increased by 0.5 kg VS·m⁻³·d⁻¹. The procedure was repeated on average every 12 days until the loading reached the level of 3 kg VS·m⁻³·d⁻¹, when the disturbance of process stability occurred, which was reflected in the increase in concentrations of VFA and alcohols, sCOD, and FOS/TAC ratio.

Increasing the OLR resulted in shortening HRT. At the first point of stability disturbance, it was 47 days for both mesophilic and thermophilic conditions (Figure 2). The process carried out in the former conditions stabilized again, while in terms of the one in the latter conditions, an imbalance occurred. According to the literature, the time that is needed for the multiplication of microorganisms in the process of methane fermentation in thermophilic conditions is between 5 and 10 days, whereas in mesophilic conditions, the multiplication occurs after between 10 and 20 days, thus it can be assumed that the reason for the process disturbance was not a too-short retention time [64,65,66].

After the next stabilization of the gas production process in mesophilic conditions, further attempts to increase OLR were made. This resulted in an insignificant increase in the production of methane and biogas.

3.3. Changes in the pH, FOS/TAC, VFA, Alcohols, and Ammonia

The analysis of changes in pH in bioreactors was conducted daily, while the FOS/TAC ratio and concentrations of VFA, alcohols, and ammonia were measured before every OLR increase and on days when process disturbances occurred (Figure 3, Figure 4 and Figure 5).

pH values during most of the experiment course in both types of conditions were at a relatively constant level of 7.3–7.6 (Figure 3). They started to rapidly decrease only in the final phase of the experiment, when the OLR value was 4 kg VS·m⁻³·d⁻¹ for mesophilic conditions and 3 kg VS·m⁻³·d⁻¹ for thermophilic conditions. At that time, pH values were 6.09 and 5.49, respectively. It is crucial to add that the concentration of ammonium as a factor that increased the pH of the solution was at a stable level of about 1 g∙dm⁻³ (Figure 4).

In addition to pH, which is the basic parameter measured in biogas installations in order to control the stability of the process, it is necessary to analyze the ratio of volatile fatty acids and total alkalinity (FOS/TAC). The value of this ratio can be an indicator of a fermentation chamber load. As confirmed in studies conducted by Liu et al., 2009, values of FOS/TAC in the range between 0.3 and 0.4 reflect a stable process [67]. Values below 0.3 indicate a possibility to increase the bioreactor load, whereas values above 0.4 indicate overloading.

Total alkalinity (TAC) in the initial phases of the experiment was 30% higher in mesophilic conditions, but on day 51, after replacing almost half of the bioreactor volume, the parameter value reached the same level of 6200 mg CaCO₃·dm⁻³ for both types of conditions.

Further increasing OLR (up to the level of 2.5 kg VS·m⁻³·d⁻¹) did not significantly influence the TAC parameter but resulted in a slight increase in FOS at a temperature of 52 °C (2209 mg) in comparison to mesophilic conditions (1728 mg). This was reflected in the FOS/TAC ratio, which reached the value of 0.36 and 0.28 in thermophilic and mesophilic conditions, respectively. Other observations at that time included increased concentrations of methanol and acetic acid, which can suggest a lack of cobalt as a key element in the metabolic pathway of methane from methanol [68]. Additionally, studies by Park et al. show that in the case of increased concentration of partial hydrogen in the fermenter, methanogenesis, in which methanol is used as the main source of carbon, is the most energetically efficient [69].

Increasing the loading to 3 kg of the OLR caused an increase in the concentration of FOS from 1700 mg CH₃COOH∙d⁻¹ to 3450 mg in mesophilic conditions and from 2200 mg to about 6000 mg in thermophilic conditions on day 83. The main identified metabolic intermediates at that time were methanol and acetic acid, as well as propionic acid and ethanol, but in much lower concentrations. In thermophilic conditions, it was observed that the concentration of methane decreased, while the FOS/TAC ratio increased from 0.3 to 0.8 within 7 days.

While analyzing changes in the range of metabolic intermediates in thermophilic conditions, an interesting relationship was observed; namely, the presence of methanol (0.339 g∙dm⁻³) in the reactor was detected even when the OLR value was 2 kg VS·m⁻³·d⁻¹ (Figure 5). Additionally, the concentration of acetic acid increased from 0.150–0.338 g∙dm⁻³ to 0.447 g∙dm⁻³. Increasing the loading to 2.5 kg VS·m⁻³·d⁻¹ resulted in a further increase in the concentration of methanol (up to 0.579 g∙dm⁻³) and acetic acid (up to 0.722 g∙dm⁻³) and a 20-fold increase in the concentration of propionic acid (to 0.242 g∙dm⁻³). The FOS/TAC ratio at the time was in the range between 0.26 and 0.36, which indicates a stable process.

The further increase in the load up to 3 kg VS·m⁻³·d⁻¹ resulted in the 5-fold, 3-fold, and 4-fold increase in the concentration of methanol (up to 2.757 g∙dm⁻³), acetic acid (up to 2.194 g∙dm⁻³), and propionic acid (up to 1.057 g∙dm⁻³), respectively. Further feeding at the same level of the load led to a linear increase in the concentrations of aforementioned metabolites until the concentration of propionic acid was above 1.7 g dm⁻¹. Since that moment, the synthesis of acetic acid was rapidly suppressed, and further observations revealed the presence of ethanol in the reactor, as well as the exponential increase in the concentration of methanol. When the concentration of methanol reached the level of about 13 g dm⁻¹, its synthesis was suppressed towards the synthesis of butyric acid, whose concentration in 13 days increased up to over 3.5 g dm⁻¹. At that time, the pH value in the reactor decreased from 7.00 to 5.49, and the FOS/TAC ratio increased from 1.85 to 9.72.

In terms of mesophilic conditions, it was observed that the process stabilized again, which was probably caused by the adaptation of microflora to an increased load, when OLR was equal to 3 kg. A similar phenomenon was noticed by [54,59]. It is worth mentioning that even though there were significant increases in the FOS/TAC ratio, no crucial changes in pH values were observed.

For most of the experiment course in mesophilic conditions, the values of sCOD were at a level below 3.5 g∙dm⁻³. In the period of destabilization, the concentration increased to 5.64 g∙dm⁻³, and after stabilizing the process, it remained at the elevated level of about 5.00 g∙dm⁻³ for dozens of days, whereas in the last phase of the experiment, when a rapid decrease in pH and biogas production was observed, it reached the value of 15.32 g∙dm⁻³. It was decided to terminate the experiment at that time. The observed increase in sCOD resulted from the accumulation of intermediate metabolites, as evidenced by chromatographic analysis and elevated FOS concentrations.

The main metabolic intermediates that were included in sCOD were methanol and acetic acid. The maximal concentration of methanol was 1.687 g dm⁻³ in the destabilization phase, while the concentration of the latter was 1.532 g∙dm⁻³. Other metabolic intermediates that were present in elevated concentrations were ethanol (0.736 g∙dm⁻³) and propionic acid (0.581 g∙dm⁻³).

In terms of the acidified reactor, methanol and acetic acid were also the main components with concentrations of 4.782 g dm⁻³ and 4.467 g∙dm⁻³, respectively. Other components that were present in lower concentrations were propionic acid (1.469 g∙dm⁻³), butyric acid (1.221 g∙dm⁻³), and ethanol (0.942 g∙dm⁻³).

3.4. Yield of Biogas per Mg of VS

The yield of biogas per Mg of VS in both mesophilic and thermophilic conditions was the highest for the load at the levels of 0.5 and 1.0 kg VS·m⁻³·d⁻¹ (Figure 6). This is due to the supply of small amounts of the substrate, which are not enough to satisfy the needs of the microflora present in the bioreactor. Because of the above, the substrate was fully utilized. As the loading increases, microorganisms use primarily the most easily decomposable compounds. Moreover, the value of HRT gets lower, which means that most of the undecomposed substrate is taken away from the reactor during the feeding process. In terms of higher values of OLR, differences in the biogas yield were significantly smaller, i.e., from about 4–9% for the yield at a loading of 1.5 kg VS·m⁻³·d⁻¹. Increasing the loading to 3.5 and 4.0 kg VS·m⁻³·d⁻¹ in mesophilic conditions and to 3.0 kg VS·m⁻³·d⁻¹ in thermophilic conditions resulted in a decrease in the methane yield by 24.79%, 55.85%, and 20.35%, respectively.

For all levels of the load in a stable period of fermentation, the process in thermophilic conditions revealed a greater yield of methane, i.e., from 0.63 to 5.48%, in comparison to the process conducted in mesophilic conditions.

4. Discussion

The performance of anaerobic digestion is determined by a complex interplay of substrate characteristics, operational parameters, and microbial community dynamics. In the present study, particular emphasis is placed on nutrient balance, organic loading rate (OLR), and temperature regime, as these factors strongly influence both the stability and efficiency of methane production.

The C/N ratio of the fruit- and vegetable-based food waste used in this study was 20.66, which falls within the optimal range of 20–30 reported in the literature for efficient methane fermentation [58,70]. Such a balanced nutrient ratio supports stable microbial activity, minimizing the risk of ammonia inhibition or nitrogen deficiency. This kind of substrate contains a relatively low amount of fat [71]. On the one hand, it is a drawback due to its low energy density; however, it does not mean that such installations are not economically affordable, for instance, at food and vegetable processing plants. Moreover, a low amount of fat eliminates the hazard of inhibition caused by long-chain fatty acids, in contrast to food waste that consists of kitchen and restaurant food waste [57].

When the OLR level reached 4 kg, it was observed that the volume of produced gases started to gradually decrease due to the inhibition of metabolic activity of microflora that occurred because of the toxic concentration of metabolic intermediates until the moment of experiment termination on day 154 [58,72,73,74]. Despite the fact that the multiplication time of microflora in mesophilic conditions is longer, the HRT observed on day 35 was still much higher than the critical value of about 20 days [64,75]. Consequently, the use of food waste as a substrate does not reduce the hydraulic retention time below the critical threshold of approximately 15 days, which could result in the acidification of reactors due to the process of leaching methanogens from the reaction environment [64,76].

Differences in the fermentation dynamics in mesophilic and thermophilic conditions were mostly caused by different reaction rates. In mesophilic conditions, the major pathway of methane synthesis is the acetotrophic one, whereas in thermophilic conditions, the main pathways are the hydrogenotrophic and methylotrophic ones [77,78]. It is also supposed that an increase in temperature results in increases in hydrolysis and acidogenesis rates, while the rate of methanogenesis is slower [49,57]. This explains the process disturbance in mesophilic conditions and the fast process collapse in thermophilic ones. Food waste, as a substrate that easily decomposes, was very quickly converted into VFA and alcohols. While in mesophilic conditions, such metabolites were almost immediately converted into methane, in thermophilic conditions, the rates of hydrolysis and acidogenesis were much faster than the rate of methanogenesis.

This also explains the higher yield of methane in thermophilic conditions. The increase in the substrate decomposition rate led to converting more substrate into methane than at the temperature of 39 °C. That is why, when reactor loadings were the lowest, because the substrate supply was lower than the minimal demand of microflora, the difference in the methane yield for both types of conditions was very low (0.63%), while when loads were higher, the difference reached the level of 5.48%. Similar results were found by Zhang et al., whose studies proved that the highest yields of fermentation of sewage sludge and food waste were at the OLR level of 3.5 kg [66]. In contrast, Yirong, in her doctoral dissertation, did not prove any differences in the methane yield from food waste in mesophilic and thermophilic conditions [79].

The most common reason for the collapse of food waste fermentation is the inhibition of VFA degradation caused by an elevated level of ammonia [57]. In the discussed study, the concentration of ammonia was at a stable, safe level throughout the course of the experiment. This was related to the fact that the substrate did not include high-protein components, which are precursors of ammonia. Thus, the process collapse needed to be caused by other factors.

The disturbance of process stability in mesophilic conditions was mainly connected to the increase in concentrations of methanol and acetic acid in the fermenter. This is contradictory to the results published by Molino et al. [80], which state that the course of food waste fermentation does not include acidogenesis. The increase in the concentrations of propionic acid and ethanol was also noted at that time. The presence of acetic and propionic acids as main metabolic intermediates was also proven by Appels et al. and Zhang et al. [49,58]. Methanol and ethanol were not mentioned in the above studies since they were not analyzed by the authors. Works by Ye et al. [81] state that the main metabolite intermediates in properly functioning bioreactors are acetic and butyric acids, whereas propionic and acetic acids are mostly present while the process is not stable. This is not in line with the results obtained during the discussed study, because butyric acid was present only in acidified reactors, both in mesophilic and thermophilic conditions. Such relationships were also noted by Appels et al., Pontoni et al., and Bong et al. [49,57,82].

While observing the process disturbance in thermophilic conditions, it was noted that in the first phase, there was a rapid increase in the concentrations of methanol and acetic acid, which resulted in many other changes. According to various literature data, accumulating acetic acid even at a concentration of 0.8 to 1.8 g∙dm⁻³ causes the inhibition of bacteria converting propionic acid to acetate [57,83,84]. This is a probable cause of the rapid increase in the concentration of propionic acid, up to the level of 3.238 g∙dm⁻³, when the concentration of acetic acid was above 3.5 g∙dm⁻³. As a consequence, accumulated propionic acid, which is toxic at concentrations above 0.8 g∙dm⁻³ [58,71], caused the inhibition of methanogenesis, both in the pathway of conversion of methanol to methane, as well as acetic acid [45].

The synthesis of methanol was inhibited beginning from day 92, and the substrate decomposition metabolic pathway turned into the synthesis of butyric acid, during which hydrogen is formed [85]. The process occurred with the rapid decrease in pH values, from 7.05 to 5.49 within 9 days. The presence of hydrogen, whose concentration in biogas was over 1000 ppm (maximum measuring range of the gas analyzer), was noted at that time. Moreover, Ahring and Westermann proved that acetic acid at a concentration of about ok. 4.9 g∙dm⁻³ has a toxic influence on microorganisms degrading butyrate [84], which could also be caused by a rapid increase in the concentration of hydrogen [84,86].

A central limitation of the present study is the absence of microbiological analyses to identify and characterize the microbial community responsible for the observed process dynamics. Such data would provide deeper insight into the mechanisms underlying process stability and inhibition. Due to budgetary and logistical constraints, microbiological evidence could not be obtained; therefore, some interpretations remain hypothetical. Further research should prioritize microbial community profiling to validate and expand upon these findings.

5. Conclusions

Food waste is a substrate that is characterized by large energy potential. Due to its characteristics, it can be used as a monosubstrate. However, increasing the reactor load needs to be done gradually to allow for achieving proper amounts of particular groups of microorganisms in a fermenter (especially in terms of methanogens, which are characterized by the longest generation time). It is also needed to control trace elements, whose deficiency significantly influences the process of destabilization. Special attention must be paid to the process conducted in thermophilic conditions, which requires the systematic control of concentrations of metabolic intermediates. Otherwise, the number of interconnected reactions can result in the collapse of the methane production process. When the process startup is properly conducted, fermentation in thermophilic conditions can result in greater methane yields produced from the same substrate amount. Future studies should include detailed identification and characterization of the microbial community involved in the process, as this would provide a more precise understanding of the observed dynamics and inhibition phenomena.