Anaerobic Digestion of Poultry Droppings in Semi-Continuous Mode and Effect of Their Co-Digestion with Physico-Chemical Sludge on Methane Yield

Abstract

Poultry waste is rich in organic matter, which allows its treatment by the process of anaerobic digestion (AD) to reinforce economic and environmental green strategies. The aim of this study is to assess the technical feasibility of poultry waste AD in semi-continuous mode and to intensify methane production by co-digestion with physico-chemical sludge, i.e., the product of the primary treatment of wastewater after the slaughtering process. First, the AD of poultry droppings is conducted in a continuous stirred anaerobic digester (CSAD) at 37 °C. A volume of 0.791 and 0.623 Nm³·Nm⁻³ reactor·days⁻¹ of biogas and methane, respectively, were produced during the entire process stabilization period. Biochemical characterization of the substrate and the final digestate show high reduction rates of the biochemical fractions, which corroborates with methane yield. In addition, the co-digestion of poultry droppings with two different ratios of physico-chemical sludge (20% and 40% added sludge considering the amount of volatile solids) shows an increase in methane production versus droppings alone, with a higher increase of 54% with a sludge ratio of 40%. Altogether, these promising results were obtained in stable processes, highlighting the pertinence of our study.

1. Introduction

Poultry is the second most consumed meat worldwide. To meet the growing demand, global poultry meat production increased from 9 to 132 million tons between 1961 and 2019, and egg production went from 15 to 90 million tons for the same period. According to the FAO, 2022, poultry meat accounted for about 39% of global meat production in 2019. Therefore, tremendous quantities of poultry production waste were generated over the past decade. For instance, in France in 2008, the total production of poultry waste was estimated at about 5.6 million tons/per year, including 2.5 million tons of manure, 2.5 million tons of slurry and 0.6 million tons of droppings per year [1].

Poultry droppings are very rich in important fertilizing mineral elements. Particularly, the large amounts of nitrogen and phosphorus favor the fertilization of the soil and improve its biological and physico-chemical properties, making it more resistant to aggressions such as drought, diseases and toxicity. Trace elements such as copper, zinc or magnesium contribute as well to these reactions. In addition, these droppings account for sources of organic matter that allow for their use as an amendment for soil regeneration [2,3].

On another note, the demand for renewable and biobased energy, coupled with the concern for livestock waste management, has fueled interest in methane production from livestock waste [4,5]. Since poultry droppings are rich in organic matter, they are considered a potential substrate for biogas production through AD [6]. AD is a biological process involving degradation of organic matter and is based on natural fermentation reactions carried out by populations of anaerobic bacteria. It leads to the formation of biogas, which is a flammable mixture composed mainly of methane and carbon dioxide in a ratio of about 3:1 by volume [7].

On a laboratory scale, the AD process can be applicable in batch, semi-continuous or continuous mode. In our previous study, we assessed the AD process in batch mode to determine the methanogenic potential (BMP) of different poultry waste, such as viscera, droppings, sieving and sludges recovered from wastewater treatment after the slaughtering process. The highest BMP was obtained with substrates rich in fat (such as viscera and sieve potential). Since poultry waste is known for its high nitrogen content, which is a potential inhibitor of the process, we evaluated the effect of ammonia concentration on methane yield and assumed that the inoculum used (LIGER) had the capacity to tolerate high ammonia concentration since no inhibition was observed at [N-NH₃] = 6.06 g·L⁻¹. In addition, the influence of the nature of the inoculum and the substrate/inoculum ratio on the methane potential and kinetics of the AD of poultry waste was reviewed [8].

In the current study, we mainly evaluated the AD of poultry waste in semi-continuous mode. CSTR is a continuous stirred tank reactor recommended for AD processes because of its simple configuration, easy operation, effective uniform stirring and proper temperature and pH maintenance. In these types of bioreactors, the biogas-generating microbial population is suspended in the digester through intermittent or continuous mixing by means of an agitator, providing good substrate–sludge contact [9,10,11]. However, when the reactor is fed with an input characterized by high total solids (TS), problems related to lab pumps being clogged by solids or the sedimentation of solids in the tubes may occur. In this case, the bioreactor cannot be fed constantly. It will thus be fed periodically, and we refer to these configurations as CSAD [12]. The CSAD and the CSTR are thus both continuous stirred anaerobic digestors, and they differ only in terms of feeding mode.

The feeding of AD units can be performed with a single substrate or with a mixture of two or more substrates (co-digestion) [13]. Co-digestion is essential to prevent certain problems, such as a lack of nitrogen, excess nitrogen, the risk of acidification due to high biodegradability, a lack of microelements and the presence of long-chain fatty acids (LCFAs). It also offers a wide range of benefits in comparison to mono-digestion, such as dilution of toxic substances, nutrient balance, enhancement of microorganism activity and subsequent methane yield intensification [14]. For example, co-digestion of poultry manure with agricultural waste resulted in 32% higher methane yields alongside an ammonia accumulation reduction of 43.7%, therefore reducing the toxicity commonly encountered in avian manures alone [15,16]. This method of optimization is economically feasible and has other advantages, such as valorization of an increased load of easily biodegradable organic matter, dilution of toxic substances, the improved buffering capacity of the mixture and the quality of the digested product [17,18,19].

As mentioned earlier, the mono-digestion of poultry droppings can particularly encounter some inhibition problems due to their nitrogen richness [5]. When exceeding a certain threshold (1500 and 7000 mg·L⁻¹), the latter can inhibit the AD process [8]. Therefore, co-digestion is recommended [17]; however, the co-substrate should be carefully chosen whilst taking into account its composition and effect on the main substrate [20], and the mixture ratio should also be optimized [21]. Poultry droppings can be successfully treated with whey, fruit and vegetable waste, organic fractions of municipal solid waste, rice straw, digested sludge and even other types of manure, including cattle slurry, buffalo manure and sheep manure [18,22]. In our previous study, the physico-chemical sludge (PCS), i.e., the product of the primary treatment of wastewater after the slaughtering process, showed a high methane yield that was higher than the yield of the droppings. Moreover, it turned out to be easily degradable, and its recovery and preparation are not complex compared to other poultry slaughterhouse wastes; after preparation, a very homogeneous sludge can be obtained [8]. For these reasons, PCS was chosen here as a co-substrate.

Given the aforementioned considerations, this paper focuses on the assessment of methane production by AD of organic laying hen droppings. The main objectives were (1) to evaluate the biogas and methane production of the droppings in a scaled-up process in semi-continuous mode using a continuously stirred anaerobic digestor (CSAD), (2) to characterize the biochemical content impacting the process, and (3) to investigate the effect of co-digestion with PCS on methane yield. The novelty of our study lies in the exploration of the biochemical content of poultry droppings, which is rarely studied for this type of waste, and relating its reduction to the obtained methane yield; further novelty is present in the method used for process optimization, which involves co-digestion with another product of this avian sector (PCS) instead of the use of other wastes (fruit and vegetable waste, municipal waste) commonly used in other studies. Altogether, the data obtained are useful for conducting an industrial technical evaluation of the digester and the process.

2. Materials and Methods

2.1. Substrate and Inoculum Preparation

Two substrates were used in our study: organic laying hen droppings (OLHDs) provided from EPLEFPA Le Gros Chêne Pontivy-France, and PCS from the slaughterhouse wastewater treatment plant of Les Volailles de Keranna Guiscriff-France. The collection of OLHDs was made using a conveyor belt divided into 3 parts that was placed under the laying hens. Every day, a part of the belt was emptied so that the droppings were not more than 3 days old. The preparation of OLHDs consisted of dilution with tap water (dilution factor 1/3), mixing and homogenization in a blender, and sieving using a 3.15 mm sieve, with droppings subsequently stored at −20 °C until use. The PCS was only homogenized and stored at −20 °C.

Two active sludge inocula were used in our study. The first was recovered from the LIGER renewable energy center in Locminé, France (a territorial facility that treats agrifood waste and animal waste), and the second was obtained from EPLEFPA Le Gros Chêne Pontivy-France (animal waste treatment plant). LIGER inoculum was sieved using a 1 mm sieve and then kept at room temperature for a fasting period of almost a week prior to the experiment. The purpose of this step was to eliminate a large amount of residual biodegradable substrates and to facilitate the evaluation of the substrate’s biodegradability [23,24]. The second inoculum was sieved using a 1 mm sieve and then a 0.4 mm sieve before being directly used in experiments without the fasting period to accelerate the start-up phase of the continuous reactor [24].

2.2. Physico-Chemical Analyses

In order to determine total solids (TS), the substrate and inoculum samples were placed in an oven at 105 °C (model 600, Memmert, Schwabach, Germany) for about 3 days until a constant weight was reached according to the NF ISO 11465 standard [25]. Subsequently, the determination of volatile solids (VS) was performed in a muffle furnace for 2 h at 550 °C based on the NT U 44-160 standard [8,23,26]. Each analysis was performed in triplicate.

2.3. Van Soest Fractionation

The samples used for this analysis had to be dried and crushed to obtain particles of almost 0.2 mm. Three detergents were then used:

-

Neutral detergent solution (NDS): disodium phosphate (4.5 g∙L⁻¹), sodium tetraborate (6.81 g∙L⁻¹), α-amylase (0.25 g∙L⁻¹), sodium EDTA (18.6 g∙L⁻¹), sodium lauryl sulfate (30 g∙L⁻¹) and sodium sulfite (5 g∙L⁻¹).

-

Acidic detergent solution (ADS): 20 g∙L⁻¹ trimethylketylammonium bromide and 98 g∙L⁻¹ H₂SO₄.

-

Sulfuric acid solution (ASS): 1317 g∙L⁻¹ of sulfuric acid.

Figure 1 represents Van Soest fractionation and the fractions obtained at each step. The first step of this method is to extract the soluble fraction (containing nonstructural carbohydrates, pectins, mucilages, soluble tannins at neutral pH, lipids and soluble proteins) with an excess of NDS, acting on 1 g of the dry sample for 1 h at 100 °C. The extracted fraction is then separated from the neutral detergent insoluble fibers (NDR, neutral detergent residue) by filtration. Sequential to the action of the neutral detergent, excess ADS acting on the NDR for 1 h at 100 °C permits the extraction of hemicellulose. The extracted fraction is separated from the acid detergent insoluble fibers (ADR, acid detergent residue) by filtration. As for cellulose, it is extracted by treating the ADR fraction with ASS for 3 h at room temperature. The filtration residue, named ASSR, corresponds to the lignin associated with inorganic elements. Hemicellulose and cellulose contents were calculated as the difference between NDR and ADR fractions and ADR and ASSR fractions, respectively [23,27]. All products used in this method were supplied by Merck, Germany. For each substrate, three samples were analyzed.

2.4. Lipid Determination (Soxhlet Method)

The Soxhlet method was used for the extraction of lipids with a solvent according to the AOAC 2003.05 standard [28]. The solvent used in our case was petroleum ether. Extraction of each substrate was performed in triplicate. A total of 2 g of dry sample was weighed and placed in a cellulose cartridge. Subsequently, 90 mL of petroleum ether (solvent provided by Lab-Scan Ltd., Unit T26 Stillorgan Ind. Park, Co. Dublin, Ireland) was poured into a previously weighed glass crucible, to which some glass beads were introduced. The cartridge and the crucible were then placed in the extractor. During the extraction, the boiled solvent (boiling temperature 40–60 °C) was evaporated, condensed and collected in the glass Soxhlet chamber. The solvent then filled the main chamber of the extractor, dissolving some of the lipids in the sample. Once the chamber was nearly full, it was emptied through the siphon, returning the solvent to the crucible to repeat the process. The extraction took 6 h to evaporate all the solvent. After extraction, the crucible containing the lipids was placed in an oven for 30 min at 105 °C and was then placed in the desiccator for 30 min to cool down. Finally, the crucible was weighed and the amount of lipids was determined by the following equation (Equation (1)):

$$\% \mathrm{lipid}=\frac{\mathrm{m}2-\mathrm{m}1 }{\mathrm{mE}}\times 100$$

(1)

where m1 = empty crucible, m2 = crucible with sample and mE = weight of the sample.

2.5. First Experiment of AD in Semi-Continuous Mode (Experiment 1, E1)

2.5.1. Description and Operating Principle of the CSAD Reactor

The CSAD, shown in Figure 2, is a perfectly mixed reactor in a tubular vertical glass form. It is equipped with peristaltic pumps that maintain the homogenization of the digester’s content by recirculation of the biogas or liquid, which also allows for the correct maintenance of temperature and pH. Note that its useful volume displays 12 L. The tool is automated using a programmable controller from Perrax (Perrax, Union, France), which collects data on operating conditions, modifies the program if necessary and automatically starts the peristaltic pumps. The hot water circulating in the digester jacket provides temperature control at 37 °C [29]. In order to maintain the same volume in the reactor and a stable hydraulic residence time (HRT), a precise mass of digestate was withdrawn, and the same mass of OLHDs was manually introduced into the reactor via the same valve using a 100 mL plastic syringe. A non-return valve is essential during manual injections and extractions to avoid any disturbances in biogas measurement that could occur.

Biogas and methane production was measured with a gas meter (Type MGC-1 PMMA, Ritter, Germany) presenting a range between 1 mL and 1 L per hour. The inlet of the first biogas meter (1) was connected directly to the reactor. Its output was connected to a bottle of 3M NaOH solution (2) for CO₂ capture, with the output also connected to another meter of the same type (3) to measure the volume of methane produced.

2.5.2. Experimental Conditions: Methane Production by Poultry Droppings in the First Experiment (E1)

OLHDs, supplied by the EPLEFPA Le Gros Chêne Pontivy-France, were used as a substrate in this semi-continuous AD experiment. Grinding and sieving were necessary to avoid the pilot pipes becoming clogged with fibrous material from the droppings.

The digester was operated under mesophilic conditions at a temperature of 37 °C. Feeding/withdrawing operations were performed daily with an HRT of 30 days and an organic loading of 2.301 kg VS∙m⁻³∙d⁻¹. The applied organic loading rate (or OLR) is calculated by the following equation (Equation (2)) [30]:

$$\mathrm{OLR}=\frac{{\mathrm{S}}_0}{\mathrm{HRT}}={ \mathrm{S}}_0\times \frac{\mathrm{Q}}{\mathrm{V}}$$

(2)

where S₀ denotes the concentration of the injected substrate (g∙L⁻¹, based on VS), Q signifies the feed rate (L∙day⁻¹) and V represents the reactor volume (L).

The feeding and withdrawal rates in the digester are related to the applied HRT and are calculated according to the following equation (Equation (3)):

$$\mathrm{Feed} \mathrm{rate}=\frac{\mathrm{V} \mathrm{reactor}}{\mathrm{HRT}}$$

(3)

where the feed rate is expressed in (L∙day⁻¹), the useful volume of the digester (Vreactor) is expressed in L and HRT is expressed in days.

To maintain a stable HRT of 30 days throughout digestion, 400 mL of digestate was then discharged daily, and the same volume of OLHDs was introduced manually each day into the reactor via a sampling valve and the use of a plastic syringe (100 mL).

Initially, the reactor was filled with 12 L of the LIGER inoculum, chosen mainly for its stable behavior [8]. The feeding started with an OLR of 0.575 kgVS·m⁻³ reactor·d⁻¹, with OLR then gradually increased (start-up phase) until a load of 2.30 kgVS·m⁻³ reactor·d⁻¹ was reached. In total, 45 days of load ramping were required to reach this goal, followed by 60 days of monitoring (reflecting the acclimation period). Overall, the reactor was run for 145 days (>3 × HRT) to ensure the stability of the physico-chemical parameters and gas production as we sometimes encountered plugging and shutting down problems with the reactor. In fact, when a reactor is clogged, circulation can be stopped for 1–2 days and the actions of feeding and withdrawal also cease so that the digester can resume its normal activity [12]. The latter was recovered immediately in our case, and this is well observed in the stable results described below.

2.6. Second Experiment of AD in Semi-Continuous Mode (Experiment 2, E2)

2.6.1. Description and Operating Principle of 2 L Reactors

This measurement system consisted of four main units: The incubation unit (1) was a thermostatically controlled (37 °C) water bath containing 2 L bioreactors (Bioprocess Control, Lund, Sweden). Each bioreactor was equipped with two closable tubes: the first for digestate withdrawal and the second for substrate supply. Agitation was performed manually twice a day for one minute. Next, the produced biogas ran through a tube connecting the incubation unit to the CO₂ fixation unit (2) made with bottles containing an alkaline solution (80 mL of 3M NaOH) and thymolphthalein as a pH indicator. Finally, the volume of CH₄ released from this unit was measured with another gas meter (Type MGC-1 PMMA, Ritter, Germany), thus forming the methane measurement unit (3) [31].

2.6.2. Experimental Conditions: Evaluation of the Co-Digestion of Poultry Droppings with Physico-Chemical Sludge on Semi-Continuous Methane Production in the Second Experiment (E2)

In order to evaluate the effect of co-digestion on poultry droppings AD in semi-continuous mode, we used 2 L bottles as digesters. Six bottles were used under a temperature of 37 °C and an HRT of 30 days. Here, OLHDs were used as substrate and PCS as co-substrate.

Table 1. Characteristics of the reactors of each co-digestion duplicate used in E2.

Bottles	Ratios (%OLHD/%PCS)	HRT (Days)	Feed Rate (g∙Days−1)	OLR (gVS∙L−1∙d−1)	Mass of OLHDs Injected per Day (gVS)	Mass of PCS Injected per Day (gVS)	Total vs. Injected per Day (gVS)
1 and 2	100/0	30	66.67	2.17	4.34	0.00	4.34
3 and 4	80/20	30	66.67	2.04	3.26	0.82	4.08
5 and 6	60/40	30	66.67	1.92	2.31	1.54	3.85

displays bottle characteristics. We chose the co-digestion ratios of 80% droppings/20% sludge and 60% droppings/40% sludge on the basis of VS. Our calculations took into consideration Guillaume and Lendormi’ study from 2015 [29].

First, the six 2 L bottles were filled with fresh inoculum in order to accelerate the start-up phase. The bottles were divided into 3 duplicates containing substrate/co-substrate ratios of 100/0, 80/20 and 60/40, respectively (Table 1). The substrate and co-substrate, prepared in advance (as mentioned in Section 2.1), were mixed according to the corresponding ratios so that they could be injected into the reactors. Feeding/withdrawing operations were performed daily. To maintain a stable HRT of 30 days throughout digestion, 66.67 g of digestate was discharged daily and the same volume of substrate was introduced into the digester through the closable tubing using a 100 mL plastic syringe. The OLR injected daily was 2.17, 2.04 and 1.92 gVS∙L⁻¹∙d⁻¹ for the 100/0, 80/20 and 60/40 ratios, respectively. Organic load did not exceed 3 gVS∙L⁻¹∙d⁻¹, so the risk of inhibition due to a high OLR is low [29,32]. As a result, the addition of sludge negligently decreased the OLR of the mixture, and it remained in the range of 2 gVS∙L⁻¹∙d⁻¹. Thus, operating conditions for this manipulation are very similar to those applied in the previous part (12 L CSAD reactors). The experiment was conducted over 80 days, a necessary duration to ensure the stability of both methane production and physico-chemical parameters [33,34].

2.7. Analytical Methods for Digester Monitoring

2.7.1. pH

pH in anaerobic processes provides a clear indication of system performance and stability. For optimal substrate degradation, a pH value between 6.5 and 8.5 should be maintained [35]. Consequently, and to monitor the stability of the AD process, pH measurements were performed several times per week on triplicates of the digestate drawn from all our reactors using a Fisher Scientific accumet AE150 pH Benchtop Meter.

2.7.2. Determination of Total Nitrogen and Phosphorus

The digestate recovered during AD, which is very rich in essential minerals such as nitrogen and phosphorus, has a known use as soil fertilizer. The total nitrogen and phosphorus amounts can also reflect the stability of the process. Total nitrogen (N) and total phosphorus (P) were quantified using Spectroquant^® tests, with ranges of 10–150 mg N∙L⁻¹ and 0.5–25 mg P∙L⁻¹, respectively. These methods are analogous to EN ISO 11905-1, DIN 38405-9 and EN ISO 6878 [36,37,38]. Each substrate was prepared in triplicate. After adding the sample and the necessary reagents to the Spectroquant tube, the latter was heated for a certain period (one hour for nitrogen determination and 30 min for phosphorus) so that the reaction could take place. Once the tube had cooled down, it was placed in a photometer previously programmed for these methods. The peroxodisulfate oxidation method was used for the analysis of these two minerals. Phosphomolybdenum blue (absorbance at a wavelength of 310–330 nm) was the indicator used for total phosphorus determination and dimethyl-2,6-phenol (absorbance at a long wavelength of 290 nm) was the one used for total nitrogen. Finally, the concentration was automatically measured according to the Beer–Lambert law A = εlc. The method was selected by the barcode of the tubes or the AutoSelector for the respective tests. These analyses were performed on three samples of digestate recovered from the CSAD at three different periods (beginning, middle and end of the experiment).

2.7.3. Determination of the FOS/TAC Ratio (Free Organic Acids/Complete Alkalimetric Titration)

The FOS/TAC ratio considers acidity vs. alkalinity in the system, which allows for regular monitoring of the reactor and assessment of the risk of its acidification. In fact, alkalinity (TAC, Complete Alkalimetric Titration) measures buffering capacity in an anaerobic digester, as well as its capacity to maintain a stable pH. It must be maintained above 2 g eq. CaCO₃·L⁻¹ to avoid drastic drops in pH [35,39]. The concentration of volatile acids (TVA, total volatile acidity, or FOS, “Flüchtige Organische Säure” in German or Free Organic Acids) must be less than 2 g eq. CH₃COOH·L⁻¹ to prevent inhibition of the methanogenic bacteria [40]. Overall, the FOS/TAC ratio should be less than 0.4 [41].

We carried out FOS/TAC analysis on a triplicate of each sample of the digestates using the AT1000 automatic titrator (Hach, DOC022.98.93074, 10/2020, Edition 8, Germany). According to the method described in Hach DOC316.52.93087, we solubilized 5 mL of digestate sample (weighed as 5 g) in 150 mL of polypropylene (PP) and 50 mL of deionized water. We added a magnet bar to the beaker for homogeneous agitation. The beaker was then placed on the titrator and the FOS/TAC application was started; an initial pH measurement was performed and titration was then started. The sample was titrated with 0.1 N (0.05 mol∙L⁻¹) of sulfuric acid solution (H₂SO₄) to reach a pH of 5.0 in order to calculate the TAC value, expressed as mg∙L⁻¹ calcium carbonate (CaCO₃). The FOS value was then obtained after a second titration step between pH 5.0 and pH 4.4. It is expressed in mg∙L⁻¹ of acetic acid (CH₃COOH).

The FOS/TAC ratio is calculated empirically according to the Nordmann method:

$$TAC=\frac{A\times C_{tit}\times 50045}{V_{smp}}$$

(4)

$$FOS=\left(\left(B\times 4\times 1.66\right)-0.15\right)\times 500$$

(5)

where A represents the volume of titrant at pH 5.0 (mL), Ctit represents the concentration of titrant (eq/L), Vsmp represents the volume of the sample (5 mL for the Nordmann method) and B represents the volume of titrant by difference at pH 4.4 (mL), which is equal to the volume of titrant at pH 5.0 − volume of titrant at pH 4.4.

2.8. Analysis of Biogas Composition Using a GEOTECH 5000/QED

The Geotech BIOGAS 5000 (Equipement scientifiques—ES, Garches, France) is a portable gas monitor for measuring the precise composition of gases in biogas applications. The data collected by this tool are consistent and help us to verify the efficiency of the digestion process. When connected to the reactor, it automatically analyzes the biogas produced by the latter, displaying the exact proportion (in %) of CH₄, CO₂ and H₂S.

2.9. Evaluation of the Biochemical Components Reduction Rate of the Input and the Final Product

Once the process was found stable, we recovered the final digestate and dried it at 105 °C. Van Soest fractionation and Soxhlet analysis were carried out on the dry residues to evaluate the content of cellulosic fibers and lipids, respectively. We then calculated the respective reduction rate according to the following equation (Equation (6)):

$$Reduction rate \left(\%\right)=\frac{amount in substrate-amount in final digestate}{amount in substrate}\times 100$$

(6)

3. Results

3.1. Experiment 1: Mesophilic AD of Poultry Droppings in a 12-Liter Semi-Continuous Stirred Anaerobic Digester (CSAD)

3.1.1. Characteristics of the Substrate

We determined the amount of TS VS for the amount of raw OLHDs used in E1. We found an amount of 294.7 ± 6.7 g·kg⁻¹ and 207.1 ± 3.70 g·kg⁻¹ for TS and VS, respectively.

3.1.2. Biogas and Methane Production during E1

During the whole experimental period, daily biogas and methane production was recorded. After the start-up phase, constant daily production was noted at 0.791 Nm³ Biogas·Nm⁻³ reactor·days⁻¹ and 0.623 Nm³ CH₄·Nm⁻³ reactor·days⁻¹ (Figure 3 and

Table 2. Operating parameters and conditions of the experimental AD unit (CSAD reactor).

Reactor volume (L)	12
Period of operation (days)	145
HRT (days)	30
OLR (kg VS·m−3·days−1)	2.30
Biogas production (Nm3 biogas·Nm−3 reactor·days−1)	0.791
Methane production (Nm3 CH4·Nm−3 reactor·days−1)	0.623
Methane content (%)	79
Methane production (Nm3 CH4·g−1VS)	0.270
Efficiency compared to batch mode (%)	82

), with methane percentage measured as 79%. Moreover, when relating the production of methane to the quantity of material injected, we determined a production of 0.270 Nm³ CH₄·g⁻¹VS (Table 2). This result indicates an efficiency of 82% for the CSAD in comparison to the yield obtained in a previous experiment in batch mode (0.320 Nm³ CH₄·g⁻¹VS [8]).

3.1.3. Handling Monitoring and Stability Factors

The stability of the CSAD was assessed by monitoring pH, TAC and FOS. The digester was operated for 145 days at an average pH of 7.8 without adjustment throughout E1. TAC was constant at 19.8 g eq. CaCO₃·L⁻¹ and FOS was stabilized at a value of 4.13 g eq. CH₃COOH·L⁻¹ (

Table 3. Stability factors of the CSAD reactor. Values correspond to the means of the samples recovered at the beginning, middle and end of the process ± standard deviation of the measurement.

	Value of Digestate	Favorable Range
pH	7.8 ± 0.048	7–8.2
TAC (g eq. CaCO3·L−1)	19.84 ± 0.37	>4
FOS (g eq. CH3COOH·L−1)	4.13 ± 0.39	<2
FOS/TAC	0.209 ± 0.021	<0.4
N tot g·kg−1	4.33 ± 0.10	3.3–8.2
P tot g·kg−1	1.50 ± 0.04	0.52–1.61

), generating a FOS/TAC ratio of 0.209 < 0.4. Altogether, these values testify to the successful completion of the AD process [41,42].

Other factors allowing us to evaluate the stability of the reactor are the amount of total nitrogen and total phosphorus in the digestate. According to Table 3, we notice that our digestate contains 4.33 gN·kg⁻¹ and 1.50 gP·kg⁻¹. These two values are maintained in the favorable range, subsequently revealing the stabilization of the process and the good quality of the recovered digestate [43].

3.1.4. Determination of the Input and Final Product’s Biochemical Fractions and the Respective Reduction Rates

The Van Soest method consists of consecutive extractions using solvents of higher extracting power and separates the different biochemical fractions. To determine the lipid fraction in particular, we performed additional Soxhlet extraction. First, we analyzed OLHDs as the substrate injected in the digester, with the digestate then recovered at the end of the experiment.

As reported in Figure 4, soluble compounds represented the highest fraction in OLHDs (63.1 g·kg⁻¹ droppings), consisting of sugars, proteins, polyphenols, organic acids and oils, followed by hemicellulose (18.1 g·kg⁻¹ droppings) and cellulose (8.82 g·kg⁻¹ droppings). Regarding lipid content, OLHDs are low in fat (only 2.51%).

We observed a reduction in all the studied fractions of the digestate after 90 days of AD: 42% for soluble matter (including particularly a 52% reduction in lipids), 50% for hemicellulose and 78% for cellulose (Figure 4).

3.2. Experiment 2: Study of the Impact of the Co-Digestion of Poultry Droppings with Physico-Chemical Sludge on Methane Yield

3.2.1. Characteristics of Substrates and Inoculum

In this experiment, we considered PCS (a co-product from poultry slaughterhouses) as a co-substrate. The TS and VS characteristics of the substrate, co-substrate, the two co-digestion mixtures and the inoculum are presented in

Table 4. Determination of the amount of TS and VS for the substrate (OLHDs, organic laying hen droppings), co-substrate (PCS, physico-chemical sludge), the two co-digestion mixtures and the inoculum used in E2.

Characteristics	Diluted OLHDs	Raw OLHDs	PCS	Mixture 80 Droppings/20 Sludge	Mixture 60 Droppings/40 Sludge	Inoculum (Le Gros Chêne)
TS (g·kg−1)	91.70 ± 0.3	275.1 ± 0.9	56.41 ± 0.66	82.95	75.19	58.43 ± 0.22
VS (g·kg−1)	65.04 ± 0.21	195.1 ± 0.2	49.31 ± 0.06	61.14	57.68	37.03 ± 0.26

. TS and VS values of the OLHDs and the PCS are in agreement with those obtained in E1 and in our previous study [8]. As expected, the substrate/co-substrate mixtures show TS and VS values between those of the diluted OLHDs and the PCS.

3.2.2. Semi-Continuous Mode Experiment (2 L Bottles)

During the whole experimental period, daily methane production was recorded. As illustrated in Figure 5, we noticed a constant average daily methane production of 0.423, 0.579 and 0.652 Nm³∙m⁻³ reactor∙days⁻¹ with the 100/0, 80/20 and 60/40 mixtures, respectively. This result reflects the positive effect of co-digestion with PCS on methane yield. Indeed, the addition of 40% sludge provided the maximum CH₄ yield.

Moreover, a biogas composition analysis was performed with the GEOTECH biogas 5000/QED analyzer. According to

Table 5. Operating parameters and conditions of each reactor duplicate representing the three co-digestion ratios.

Substrate	100/0	80/20	60/40
Reactor volume (L)	2	2	2
Period of operation (days)	80	80	80
HRT (days)	30	30	30
OLR (gVS·L−1·days−1)	2.17	2.04	1.92
Methane flow rate (Nm3·m−3 reactor·days−1)	0.423	0.579	0.652
Methane production (Nm3 CH4·kg−1 VS)	0.248	0.388	0.380
CH4 (%) (biogas analyzer)	61	64	67
CO2 (%) (biogas analyzer)	39	36	33
H2S (ppm) (biogas analyzer)	293	29	20

, the methane percentages obtained by the three trials are within the favorable range cited by Usack et al., 2012 [12]. The highest is observed with the 60/40 mixture (67%), followed by the 80/20 mixture (64%) and the OLHDs alone (61%). In parallel, a decrease in hydrogen sulfide (H₂S) is observed with the incorporation of a higher quantity of sludge.

3.2.3. Handling Monitoring and Stability Factors

For process monitoring, we completed measurements of pH, TAC and FOS in the digesters. Two FOS/TAC analyses per week were performed on the digestate drawn from each reactor. Figure 6a–d present the results of pH, FOS, TAC and FOS/TAC ratio, respectively.

Table 6. Stability factors of each reactor duplicate representing the three co-digestion ratios. Values correspond to the mean ± standard deviation of the measurement.

	pH	FOS (g eq. CH3COOH·L−1)	TAC (g eq. CaCO3·L−1)	FOS/TAC
100/0	8.15 ± 0.21	4.35 ± 0.72	19.93 ± 2.16	0.218 ± 0.025
80/20	8.23 ± 0.100	4.40 ± 0.88	19.18 ± 1.87	0.227 ± 0.028
60/40	8.19 ± 0.15	4.06 ± 0.79	17.76 ± 1.05	0.227 ± 0.034

summarizes the average of each of these factors for each reactor throughout the experimental period.

Figure 6a shows a stable pH in all reactors, with values not exceeding 8.2 without adjustment throughout the experiment. Concerning acidity or FOS (Figure 6b, Table 6), we observed a small fluctuation in the values depending on the feeding of each digester, ranging from 3.5 g eq. CH₃COOH∙L⁻¹ to 5 g eq. CH₃COOH∙L⁻¹. The lowest average value of 4.06 g eq. CH₃COOH∙L⁻¹ corresponds to the 60/40 ratio co-digestion reactor. In parallel, TAC (Figure 6c, Table 6) fluctuates from 17 g eq. CaCO₃∙L⁻¹ to 23 g eq. CaCO₃∙L⁻¹ (depending on each reactor), the lowest average value (17.8 g eq. CaCO₃∙L⁻¹) being associated with the 60/40 ratio. Figure 6d depicts the variation in the FOS/TAC ratio throughout the AD, a ratio proportional to the elevations of FOS and TAC. This demonstrates that the average is 0.218 for the reactor without co-digestion and 0.227 for both co-digestion reactors (Table 6).

4. Discussion

4.1. Experiment 1: Mesophilic AD of Poultry Droppings in a 12 L Semi-Continuous Stirred Anaerobic Digester (CSAD)

In E1, the TS and VS values of the OLHDs are comparable to those obtained in the literature [44,45], as well as those recorded in our previous study where OLHDs presented 288 g·kg⁻¹ for TS and 205 g·kg⁻¹ for VS. [8]. In order to minimize inhibition risks that may occur during AD of chicken droppings, the recommended TS concentration in the feed should be ≤10% [8,46]. For this reason, OLHDs were diluted at a factor of 1/3 to obtain a final TS value of 9.82% (<10%).

Regarding the mesophilic AD of poultry waste, the HRT can vary between 12 and 40 days depending on the substrate and the organic load applied in the digester [47]. We set HRT to 30 days in this work in accordance with previous studies conducted on AD of agro-industrial waste in continuous mode [23,48].

Concerning the OLR, it is known that a high organic load enhances microorganisms’ growth and generates better biogas production. However, when the OLR exceeds reactor capacity, imbalances between the four stages of AD would be expected, causing irreversible acidification and process inhibition due to VFA accumulation [31,49]. In order to maximize biogas production while maintaining reactor stability, organic loading should be between 0.25 and 3 kg COD·m⁻³·d⁻¹ [32]. Mahdy et al., 2020, found that the AD of nitrogen-rich substrates using an OLR of 2.5 gTS·L⁻¹·d⁻¹ can boost ammonia-tolerant bacteria and enable them to work more efficiently while providing more methanogenic substrates (acetate and hydrogen) for methanogenesis [50]. These findings support our OLR choice of 2.30 kgVS·m⁻³·d⁻¹.

During E1, we obtained daily production of 0.791 Nm³ biogas·Nm⁻³ reactor·days⁻¹ and 0.623 Nm³ CH₄·Nm⁻³ reactor·days⁻¹. Biogas yield is comparable to that obtained in previous continuous AD studies conducted on chicken manure [51,52]. In our case, the biogas contained 79% methane. This percentage is in agreement with the range of yields observed in the literature, where methane content varies from 40% and above with chicken manure particularly [52,53,54] to 70–74% with animal manure in general [51,55,56]. In addition, when Usack et al. described the operating parameters of this type of reactor, they expected optimum CH₄ and CO₂ percentages between 55 and 70% and 30 and 45%, respectively [12]. Knowing that poultry droppings are very rich in nitrogen, which is essential for bacterial activity, it can be assumed that the higher the protein content, the higher the methane production [57]. Altogether, we can notice variability in biogas and methane production reflecting the effect of several factors, such as the nature of the substrate, its recuperation and preparation and the operating conditions (HRT, OLR, temperature, etc.).

In order to confirm the stability of the AD process in the CSAD, several factors were monitored. As shown in Table 3, FOS values were high compared to the favorable range (4.13 g eq. CH₃COOH·L⁻¹ > 2 g eq. CH₃COOH·L⁻¹). However, the buffering capacity (TAC) was also large (19.8 g eq. CaCO₃·L⁻¹ > 4 g eq. CaCO₃·L⁻¹) and therefore good enough to maintain stable conditions in the digesters. We suggest that the fast growth rate of acidogenic bacteria (almost 4 days) generated excessive FOS concentration. However, this did not lead to an inhibition of the process because acid production was coupled with acid degradation owing to acetogenic and methanogenic microorganisms, which thus resulted in high TAC [35,40,58]. Therefore, the pH value (7.8) and the FOS/TAC ratio (0.209 < 0.4) imply that the CSAD was operating under tolerable conditions for methane production and the balance of microorganisms [58,59]. Another stability indicator is the amount of nitrogen and phosphorus recovered from the digestate [43]. In our study, the values of N and P are not only in the favor of process stability, but they also reveal the mineral richness of the digestate, which can thus be considered as a good fertilizer.

To close this part, when comparing E1′s methane production to that obtained in batch mode [8], we note an efficiency of 82%. In fact, different studies have also assessed the efficiency of the continuous mode compared to the batch mode using similar conditions (HRT of 30 days, mesophilic conditions). They reported efficiencies ranging from 60 to 90% [48,60,61,62,63]. More particularly, a study carried out by El Achkar et al. in 2016 on grape pomace with the same 12 L pilot also recorded analogous efficiency (81%) between the two AD modes [23]. Accordingly, our work is consistent with the literature, indicating the success of the pilot scale-up.

Additionally, we determined the biochemical fractions of the substrate and tracked their reduction rates by comparing the composition of E1 inputs and outputs. Very few studies cover this aspect for poultry droppings, and the results found are generally heterogeneous. For example, Rahman et al. in 2018 observed the opposite of our results, where cellulosic content was higher than hemicellulosic content in poultry droppings [44]. In fact, the content of polysaccharide fractions can vary with the straw that makes up their bedding, which is aggregated to the droppings. This mixture depends on each livestock farm and must be carefully chosen since it directly influences the kinetics of the AD process [64]. Furthermore, our poultry droppings consisted of 2.51% lipids, which is comparable to the 2% lipid content found in the study by Afilal et al. in 2014, thus implying low fat content in droppings [65].

When we assessed the reduction rate of each fraction, 50% of hemicellulose and 78% of cellulose were considered degraded in the CSAD (Figure 4). Likewise, Tambone et al. in 2009 showed a higher reduction of cellulose (63%) than hemicellulose (58%) during AD of a mixture of energetic crops, cow slurry, agro-industrial waste and organic fractions of municipal solid waste (OFMSW) in a large-scale mill [66]. Among different components available in lignocelluloses, cellulose and hemicellulose are mostly degraded by anaerobic microorganisms during the AD process [67]. Hemicellulose is known for its easy degradation; it provides an important carbon source for anaerobic microorganisms [31,68,69]. In contrast, cellulose is formed by a crystalline structure that prevents the penetration of microorganisms or extracellular enzymes [67,70], though it can also exist in an amorphous form (non-crystalline) that is soluble and easily digested by microorganisms and enzymes [71]. Additionally, it is important to highlight the positive effect of the substrate grinding that was performed prior to the experiment, which generated a reduction in size that was necessary for an easily digestible lignocellulosic biomass that flowed well [72,73]. All these findings guarantee the high reduction rate of the biochemical fractions, which in turn explains the high methane yield obtained with poultry droppings.

4.2. Experiment 2: Study of the Impact of the Co-Digestion of Poultry Droppings with Physico-Chemical Sludge on Methane Yield

In E2, the operating conditions were very similar to those applied in E1. HRT was set at 30 days [48,61] and OLR did not exceed 3 gVS∙L⁻¹∙d⁻¹ [32] so that reactor stability could be maintained.

Our approach implied co-digestion, which is a process whereby two high-energy organic wastes are mixed together to enhance the degradation of the product of interest and thus the production of methane. Our results showed that methane yield increased with the co-substrate/substrate ratio, with a 37 and 54% increase in methane production resulting from the respective addition of 20 and 40% sludge. In fact, mixing manure with other carbon-rich organic wastes improves the nutritional balance; higher organic load can therefore be degraded, and better buffering capacity of the mixture, higher biogas yield and better quality of the digested product can be achieved [19,74,75,76]. In 2015, Abouelenien et al. examined the co-digestion of poultry manure with agricultural waste and observed the highest performance under mesophilic conditions, with a 41.3% increase in methane production for a substrate/co-substrate ratio of 7:3 (wet weight basis) [15]. Similarly, a 32% swell in methane yield was provided after the addition of fruit and vegetable waste to poultry manure (1:1 wet weight ratio), which was attributed to a reduction in the ammonia concentration commonly found in manures alone [16]. Furthermore, the co-digestion of food waste, sewage sludge and poultry litter at different ratios generated high biogas yields and particularly enhanced methane content [77]. Another study on co-digestion of pig manure with winery wastewater revealed that the addition of 40% of the co-substrate improved volatile matter removal by up to 61% and increased methane yield [78]. The choice of PCS as a co-substrate was based on our previous study, which demonstrated that the methanogenic potential of PCS (0.470 Nm³CH₄·kg⁻¹·VS) was greater than that of OLHDs (0.320 Nm³CH₄·kg⁻¹·VS) [8]. In fact, the sludge we use here underwent physico-chemical pretreatment at the slaughterhouse’s wastewater plant. This process allowed for the coagulation of particles, thus enabling the transformation of certain solutes (e.g., proteins, heavy metals) into flakes and facilitating their elimination. Accordingly, PCS has low organic matter content (TS < 5%, Table 4) and is deficient in inhibitory materials (ammonia, LCFA, etc.) [79]. Thus, mixing it with the manure will enhance biogas production.

When adding 40% sludge, biogas composition analysis performed with the GEOTECH 5000/QED tool showed a higher methane percentage (67%) and a decrease in H₂S quantity (20 ppm). H₂S is a toxic gas produced by the degradation of sulfates by sulfate-reducing bacteria during the AD process. Its dissemination in the air creates fine particles and secondary gases which are irritating and toxic for humans and animals. Overall, for protein-rich droppings, it is more advantageous to mix them with other organic substrates such as sludge to reduce H₂S emissions, purify the generated biogas and improve methane production.

Finally, and in regards to process stability, we noticed a lower accumulation of FOS when co-digestion was conducted with a 60/40 ratio. Adding more sludge prevents acidity in the AD process resulting from the use of poultry droppings [64]. Additionally, the proportional increments in TAC and FOS imply that the system has high buffering capacity [80]. This maintains process stability without a drastic drop in pH. Moreover, FOS/TAC ratio was less than 0.4 in all reactors, implying the stability of the process and the balance of acidogenic, acetogenic and methanogenic microorganisms under the selected conditions [41,59].

5. Conclusions

Poultry droppings contain a high amount of organic matter, which suggests their potential use as an alternative energy source in the AD process. We previously studied the methanogenic potential of this waste in batch mode and assessed it in semi-continuous mode in this study using a 12 L CSAD. We observed a production of 0.270 Nm³ CH₄·g⁻¹VS using the 12 L CSAD, indicating an efficiency of 82% in comparison to the yield obtained in the previous experiment using batch mode (0.320 Nm³ CH₄·g⁻¹VS). Note that the production of biogas and methane was constantly monitored during a stable process. Additionally, to optimize the AD process, we applied the co-digestion method. Mixing poultry droppings with physico-chemical sludge in a 60/40 ratio enhanced methane production by 54% without affecting process stability. This approach promises to generate higher amounts of methane and to further valorize wastes from the poultry sector.