Long-Term Pre-Treatment of Municipal Sewage Sludge with Solidified Carbon Dioxide (SCO₂)—Effect on Anaerobic Digestion Efficiency

Abstract

Studies on harnessing solidified carbon dioxide (SCO₂) for municipal sewage sludge (MSS) pre-treatment have been conducted exclusively in batch reactors. This makes it difficult to accurately assess how long-term SCO₂ treatment affects anaerobic digestion (AD) conditions and performance. The aim of our study was to evaluate the effect of long-term MSS pre-treatment with SCO₂ on AD conditions, anaerobic bacterial community, and biogas composition and yields. The presented experiments are the first studies on the effect of pre-treatment with SCO₂ on the efficiency of AD of MSS in continuous reactors. So far, the impact of the organic load rate (OLR) on the efficiency of MSS methane fermentation has not been assessed, which is also a novelty of the conducted research. The AD process was conducted in continuous-stirred, continuous-flow anaerobic with an active volume of 20 dm³. The digestion process was run at 38 ± 1 °C. The experiment was divided into two stages. Raw (non-pretreated) MSS was used in stage 1, whereas the MSS used in stage 2 was pre-treated with SCO₂. The SCO₂/MSS ratio was 1:3. Each stage was sub-divided into four variants, with different levels of the OLR ranging from 2.0 to 5.0 gCOD/dm³·day. Pre-treatment with SCO₂ was found to improve AD performance at an OLR of 3.0–4.0 gVS/dm³·day. The 3.0 gVS/dm³·day variant offered the best biogas production performance—both daily (29 ± 1.3 dm³/day) and per VS added (0.49 ± 0.02 dm³/gVS)—as well as the highest CH₄ content in the biogas (70.1 ± 1.0%). In this variant, the highest energy output effect of 187.07 ± 1.5 Wh/day was obtained. The SCO₂ pre-treatment was not found to change the pH, FOS/TAC, or the anaerobic bacterial community composition. Instead, these variables were mainly affected by the OLR. Our study shows that MSS pre-treatment with SCO₂ at a SCO₂/MSS ratio of 0.3 (by volume) significantly improves AD performance in terms of methane production and feedstock mineralization. The pre-treatment was found to have no negative effect on the long-term continuous operation of the reactor.

1. Introduction

Anaerobic digestion (AD) is a method commonly used for treating municipal sewage sludge (MSS), offering technological, environmental, and commercial benefits [1,2]. AD enables the controlled mineralization of labile organic matter, reduces putrescibility, mitigates odor nuisance, minimizes the sanitary risk of MSS, reduces MSS mass, and improves dewaterability [3]. There are two end products of AD. The first is stabilized sludge, which can be used in agriculture, environmental improvement, or land reclamation (due to its fertilizing potential). The other is biomethane, which is fed into gas pipelines, used as fuel for vehicles, or combusted for heat and electricity in cogeneration systems [4].

AD is a well-established technology, highly optimized on the basis of extensive data from experimental research and commercial deployment [5,6,7]. According to many researchers and operators, the prerequisite to improving the AD process and its performance lies in implementing effective and commercially/environmentally informed methods of MSS pre-treatment [8,9,10]. Many such methods have been identified and tested, including hydrothermal depolymerization, ultrasonic disintegration, hydrodynamic cavitation, as well as treatments using acids, bases, strong oxidizing agents, enzymes, and other biologically active substances [11,12,13]. Though pre-treatments have seen substantial improvements over time, and many have been deployed in large scale, new, competitive solutions are still being sought that would combine commercial viability with environmental benefits.

Studies, reviews, and experimental work conducted so far point to solid carbon dioxide (SCO₂) as a promising option for MSS pre-treatment [14,15]. The ability to harvest SCO₂ via the biogas conditioning/upgrading process greatly bolsters the attractiveness and potential of the technology, bringing it fully in line with material recycling, greenhouse gas reduction, climate change mitigation, and the principles of a circular economy [16,17,18]. SCO₂ can damage cell walls by way of ice crystal formation in the microbe-growing medium and in the cytoplasm [19]. This leads to the destruction of the sludge biomass and dissolution of organic matter, increasing its availability to anaerobic bacteria and improving AD performance [20]. Using SCO₂ for sludge disintegration has been shown to markedly increase soluble nutrient levels (orthophosphates (P-PO₄), ammonia nitrogen (N-NH₄)) and organic content (total organic carbon (TOC), chemical oxygen demand (COD), volatile fatty acid (VFA)) [21], improve biogas output by up to 40% and more [14], increase methane production [14,15], and provide commercial benefits if certain conditions are met [14,20].

Research so far on harnessing SCO₂ for MSS pre-treatment has focused exclusively on fine-tuning and optimizing SCO₂ doses for the best biogas/methane yields [14,21,22]. All such experiments have been performed exclusively in batch reactors. The data and results from respirometric studies are limited and mainly focused on the biogas output of the tested process variants. This alone cannot give a full and accurate picture of how long-term SCO₂ treatment affects anaerobic digestion conditions, trends in reactor physicochemical parameters, the evolution of anaerobic bacteria composition, and performance over time [20]. Studies in continuous-flow reactors can reproduce the conditions and process parameters in commercial-scale plants and improve the state of the technology—a vital step towards increasing the applicability of the process [23].

The aim of the present study was to evaluate the effect of long-term MSS pre-treatment with SCO₂ on anaerobic digestion conditions and trends in reactor physicochemical parameters, anaerobic bacterial community composition over time, biodegradability, production performance, biogas composition, and digestate profile. The experiments focused on comparative AD analysis of the SCO₂ pre-treated MSS with methane fermentation of the raw MSS. The second analyzed variable affecting AD efficiency in continuous fermentation reactors was the applied organic load rate (OLR) and the hydraulic retention time (HRT) as a related parameter.

2. Materials and Methods

2.1. Experimental Design

The experiment was divided into two stages, with different approaches to MSS pre-treatment prior to AD. Raw (non-pretreated) MSS was used in stage 1 (S1), whereas the MSS used in stage 2 (S2) was pre-treated with SCO₂ prior to the AD process. The SCO₂/MSS ratio was 1:3. SCO₂ doses were selected based on literature data [21] and our previous optimization studies in batch reactors [14]. Each stage was sub-divided into four variants with different levels of interrelated AD parameters for OLR and HRT. Variant 1 (V1): OLR = 2.0 gCOD/dm³·d, HRT = 21 days. Variant 2 (V2): OLR = 3.0 gCOD/dm³·d, HRT = 14 days. Variant 3 (V3) = 4.0 gCOD/dm³·d, HRT = 11 days. Variant 4 (V4) = 5.0 gCOD/dm³·d, HRT = 9 days.

2.2. Materials

The thickened MSS was sourced from a sludge processing line of the mechanical-biological municipal sewage treatment plant in Bialystok, Poland (53.16903 N, 23.08705 E). The plant uses a traditional activated sludge process with enhanced nutrient removal, working at a capacity of 100,000 m³ of wastewater/day. The anaerobic reactors were inoculated with anaerobic sludge (AS) from an enclosed 7300 m³ digester of the Bialystok treatment plant (T = 35 °C, OLR = 2.0 gVS/dm³·day, HRT = 21 days). The profiles of the raw MSS (S1), the SCO₂-treated MSS (S2), and the AS inoculum are presented in

Table 1. Characteristics of the sludge types used in the experiment.

Parameter	Unit	MSS (S1)	MSS (S2)	AS
pH	-
Total solids (TS)	[%]	4.96 ± 0.68	4.94 ± 0.92	2.24 ± 0.55
Volatile solids (VS)	[%TS]	84.91 ± 1.81	84.70 ± 1.98	67.22 ± 1.66
Mineral solids (MS)	[%TS]	15.09 ± 1.81	15.30 ± 1.98	32.78 ± 1.66
Total carbon (TC)	[mg/gTS]	585 ± 10	583 ± 12	330 ± 12
Total organic carbon (TOC)	[mg/gTS]	571 ± 15	568 ± 19	308 ± 9
Total nitrogen (TN)	[mg/gTS]	91.51 ± 6.43	90.22 ± 5.5	34.6 ± 3.8
C/N ratio	-	6.39 ± 1.56	6.46 ± 2.18	9.54 ± 3.16
Total phosphorus (TP)	[mg/gTS]	3.5 ± 1.1	3.4 ± 1.2	1.6 ± 0.2
Soluble COD	[mg/dm3]	115 ± 10	383 ± 14	1100 ± 16
Soluble TOC	[mg/dm3]	34.30 ± 3.21	127.96 ± 2.25	325 ± 11
Soluble P-PO43−	[mg/dm3]	60.33 ± 2.43	78.21 ± 4.16	65.23 ± 2.13
Soluble N-NH4+	[mg/dm3]	83.25 ± 10.51	263 ± 14	99.44 ± 3.18

.

The SCO₂ used in the experiments was in the form of granules 3.0 ± 1.0 mm in diameter (Sopel Ltd., Bialystok, Poland). SCO₂ was produced from liquid CO₂ using a Dryice—ICE Peletizer IP-100H (ICS, Považská Bystrica, Slovakia). This type of SCO₂ sublimates at temperatures exceeding −56.4 °C and pressures below 5.13 atm. Under atmospheric air pressure, sublimation occurs at −78.5 °C (enthalpy = 573 kJ/kg), making SCO₂ 3.3 times more efficient than water ice of the same volume. The SCO₂ is a natural, food-contact product, odorless, tasteless, non-poisonous, and non-combustible.

2.3. Research Station

MSS pre-treatment with SCO₂ was performed in glass, open, continuous-stirred reactors (JLT 6, VELP Scientifica, Milano, Italy). The reactors were first filled with raw MSS having a temperature of 20 °C MSS, which was then amended with SCO₂. The mixture was stirred at 50 rpm for 20 min. The samples were left until the SCO₂ was fully sublimated out, then, after reaching 20 °C, they were subjected to AD.

The AD process was conducted in continuous-stirred, continuous-flow anaerobic reactors (Gunt CE 702, Hamburg, Germany) (Figure 1) [24]. Total/active volume was 30 dm³/20 dm³, respectively. The digestion process was run at 38 ± 1 °C. The reactor was fitted with detectors, a temperature stabilization system, a stirring system with a four-blade mechanical agitator rotating at 100 rpm, a feedstock in/digestate out pump, and a biogas outflow meter. Feedstock input and digestate discharge occurred daily. The OLR and HRT values used depended on the variant of the experiment. Each variant was continued for three times the hydraulic retention time. Analyses were conducted every 4 days.

2.4. Analytical Methods

TS in the MSS was determined by drying at 105 °C. VS was quantified by combustion at 550 °C according to PN-EN 15935:23022-01 [25]. TC and TOC were determined by high temperature decomposition with infrared detection in a multi NC 3100 analyzer (Analytik Jena, Jena, Germany). Soluble COD, TN, N-NH₄, and P-PO₄ were determined by the spectrophotometric method after prior mineralization, using a Hach DR6000 spectrometer (Hach, Loveland, OH, USA) with the wavelength for COD set at 620 nm, for nitrogen set at 655 nm, and for phosphorus set at 700 nm. Supernatant was extracted via centrifugation of the MSS at 5000 rpm for 10 min (MPW-251, Med. Instruments, Warsaw, Poland). The potentiometric method was used to determine pH. Biogas was analyzed for composition in a DP-28BIO gas analyzer (Nanosens, Wysogotowo, Poland).

Anaerobic microbes were identified using FISH with DAPI counterstain and Image Processing and Analysis in Java (ImageJ, LOCI, University of Wisconsin, Madison, WI, USA) [26]. Examinations were performed under an epifluorescent microscope (Nikon, Tokyo, Japan). Hybridization was performed using molecular probes for EUB338 Bacteria, ARC915 Archaea, MSMX860 Methanosarcinaceae, and MX825 Methanosaeta. Samples for taxonomic analysis were collected at the end of each experimental variant.

2.5. Formulas and Calculations

The digestion coefficient (portion digested), i.e., the ratio of the organic VS load removed in the reactor to the VS load fed into the reactor, was determined using the following equation:

$${\mathsf{\eta }}_{\mathrm{F}}=\frac{{\mathrm{VS}}_{\mathrm{in}}·{\mathsf{\rho }}_{\mathrm{in}}·{\mathrm{Q}}_{\mathrm{in}}-{\mathrm{VS}}_{\mathrm{out}}·{\mathsf{\rho }}_{\mathrm{out}}·{\mathrm{Q}}_{\mathrm{out}}}{{\mathrm{VS}}_{\mathrm{in}}·{\mathsf{\rho }}_{\mathrm{in}}·{\mathrm{Q}}_{\mathrm{in}}}$$

(1)

where ${\mathsf{\eta }}_{\mathrm{F}}$—digestion coefficient, %; ${\mathrm{VS}}_{\mathrm{in}}$—concentration of organic matter in the influent, g/dm³; ${\mathrm{VS}}_{\mathrm{out}}$—concentration of organic matter in the digestate, g/dm³; ${\mathsf{\rho }}_{\mathrm{in}}$—influent density, g/cm³; ${\mathsf{\rho }}_{\mathrm{out}}$—digestate density, g/cm³; ${\mathrm{Q}}_{\mathrm{in}}$—daily volume of feedstock (in), cm³/day; ${\mathrm{Q}}_{\mathrm{out}}$—daily volume of digestate (out), cm³/day;

Biogas/CH₄ production per VS load was calculated as:

$${\mathrm{Y}}_{\mathrm{b}/{\mathrm{CH}}_4}^{{\mathrm{VS}}_{\mathrm{rem}}}=\frac{{\mathrm{V}}_{\mathrm{b}/{\mathrm{CH}}_4}}{({\mathrm{VS}}_{\mathrm{in}}·{\mathsf{\rho }}_{\mathrm{in}}·{\mathrm{Q}}_{\mathrm{in}}-{\mathrm{VS}}_{\mathrm{out}}·{\mathsf{\rho }}_{\mathrm{out}}·{\mathrm{Q}}_{\mathrm{out}})/1000}$$

(2)

Biogas/CH₄ production per VS load in the influent was calculated using the following equation:

$${\mathrm{Y}}_{\mathrm{b}/{\mathrm{CH}}_4}^{{\mathrm{VS}}_{\mathrm{in}}}=\frac{{\mathrm{V}}_{\mathrm{b}/{\mathrm{CH}}_4}}{({\mathrm{VS}}_{\mathrm{in}}·{\mathsf{\rho }}_{\mathrm{in}}·{\mathrm{Q}}_{\mathrm{in}})/1000}$$

(3)

where ${\mathrm{Y}}_{\mathrm{b}/{\mathrm{CH}}_4}^{{\mathrm{VS}}_{\mathrm{rem}}}$—biogas production per VS_rem, cm³/gVS_rem; ${\mathrm{Y}}_{\mathrm{b}/{\mathrm{CH}}_4}^{{\mathrm{VS}}_{\mathrm{in}}}$—biogas production per VS in the influent, cm³/gVS_in; ${\mathrm{V}}_{\mathrm{b}/{\mathrm{CH}}_4}$—volume of biogas/CH₄ produced per influent load, cm³/day; ${\mathrm{Q}}_{\mathrm{in}}$—single load of influent (by volume), cm³; ${\mathrm{Q}}_{\mathrm{out}}$—specific post-AD digestate out (by volume), cm³.

The energy output generated from methane production was calculated using the following equation:

$${\mathrm{E}}_{\mathrm{out}}={\mathrm{M}}_{\mathrm{MSS}}·{\mathrm{Y}}_{\mathrm{Methane}}·{\mathrm{CV}}_{\mathrm{Methane}}$$

(4)

where E_out—energy output, Wh/day; M_MSS—mass of MSS, gVS/day; Y_Methane—methane yield, dm³/gVS; CV_Methane—calorific value of methane, Wh/dm³.

2.6. Statistical Methods

The research was carried out in three replications. Statistical analysis was carried out using the Statistica 13.3 PL package (Statsoft, Inc., Tulsa, OK, USA). The Shapiro–Wilk test was used to verify the hypothesis regarding the distribution of the researched variables. ANOVA was used to establish the significance of differences between variances. Levene’s test was used to check the homogeneity of variances in groups, whereas Tukey’s HSD was used to determine the significance of differences between the variables. Results were considered significant at α = 0.05.

3. Results and Discussion

3.1. Dissolved Phase

The SCO₂ pre-treatment increased the levels of soluble COD, TOC, P-PO₄³⁻, and N-NH₄⁺ in the MSS. COD in the non-conditioned and conditioned MSS supernatants was 115 ± 10 mg/dm³ and 383 ± 14 mg/dm³, respectively. Soluble TOC also responded to the pre-treatment, rising from 34.30 ± 3.21 to 127.96 ± 2.25 mg/dm³. The MSS supernatant contained 60.33 ± 2.43 mgP-PO₄³⁻/dm³ and 83.25 ± 10.51 mgN-NH₄⁺/dm³ when no pre-treatment was used. The application of SCO₂ boosted these metrics to 83.25 ± 10.51 mg P-PO₄³⁻/dm³ and 263 ± 14 mgN-NH₄⁺/dm³. This increase was probably caused by disruption of MSS structures and the subsequent degradation of individual microbe cells [27]. Research has shown that the SCO₂-induced damage of MSS cell structures results in increased concentrations of soluble COD, ammonia nitrogen, orthophosphates, proteins, and molecular material in the supernatant [14]. The increased ammonia nitrogen and orthophosphate levels may also be attributable to the decomposition of nitrogen and phosphorous organics by the hydrolytic action of enzymes (contained in the microbial protoplasts and released upon the destruction of microbial cells structures by SCO₂) [28].

Zawieja (2018) [15] found similarly significant increases in COD and TOC concentrations in the supernatant after SCO₂ pre-treatment. The non-pretreated MSS contained 110 mgO₂/dm³ COD and 26 mg/dm³ TOC, compared to 296 mgO₂/dm³ COD and 78 mg/dm³ TOC determined after pre-treatment at an SCO₂/MSS ratio of 0.35/1.0 [15]. In another study, the same author (2019) [22] also observed increased N-NH₄⁺ concentration in the supernatant—from 43 mg/dm³ in raw MSS to 102 mg/dm³ at a SCO₂/MSS ratio of 0.75/1.0 [22]. Machnicka et al. (2019) [29] corroborated these findings, obtaining 63 mgO₂/dm³ COD in non-treated MSS, which rose to 205 mgO₂/dm³ after pre-treatment with SCO₂ at a SCO₂/MSS ratio of 0.25/1 and to 889 mgO₂/dm³ at a SCO₂/MSS ratio of 1:1 [29].

3.2. Biogas and Methane

Stage 2 (S2) performed significantly better in terms of biogas production, with the exact yields varying from variant to variant—from 20.7 ± 0.6 dm³/day (V1) to 29.1 ± 1.3 dm³/day (V2). In S1, the daily biogas yields peaked at 25.3 ± 1.3 dm³/day (V2) (Figure 2a), with the other variants producing 18.8 ± 0.9 m³/day (S1V1) and approx. 23.0 dm³/day (S1V3 and S1V4) (Figure 2a). Yields across all OLR values tested (V1–V4) in S2 averaged 25.1 ± 1.0 dm³/day, which was significantly higher than the average determined for S1 (22.6 ± 0.9 dm³/day) (2b). A similar pattern emerged for CH₄. The methane fraction content peaked at 70.1 ± 1.0% in S2V2, exceeding the other experimental variants by a significant margin (Figure 3a). In S2, the CH₄ content of the biogas ranged from 59.1 ± 2.2% (V4) to 65.9 ± 0.6% (V1) (Figure 3a), whereas the range for S1 was 57.4 ± 0.9% (V4) to 65.3 ± 0.6% (V1). The SCO₂ pre-treatment significantly improved the biogas composition. Average CH₄ content across the OLRs tested in S2 was 64.9 ± 1.4%, whereas in S1 it was at 61.8 ± 0.8% (Figure 2a). Daily CH₄ production peaked at 20.4 ± 1.1 dm³ in S2V2, whereas the lowest yield wase noted at 12.4 ± 0.6 dm³/day for S1V1 (Figure 3b).

Daily biogas and CH₄ yields per VS_in and VS_out significantly decreased as the OLR increased. In S1, biogas yields varied from between 0.47 ± 0.02 dm³/gVS and 0.74 ± 0.04 dm³/gVS_removed (V1) to 0.20 ± 0.01 dm³/gVS and 0.45 ± 0.01 cm³/gVS_removed (V4) (Figure 4a). The range for S2 was from 0.52 ± 0.01 dm³/gVS and 0.19 ± 0.01 cm³/gVS_removed (V1) to 0.77 ± 0.02 dm³/gVS and 0.39 ± 0.01dm³/gVS_removed (V4) (Figure 4b). In terms of the CH₄ fraction produced, the most pronounced differences were noted in V2, with 0.52 ± 0.03 cm³/gVS_removed in S2V2 and 0.42 ± 0.02 cm³/gVS_removed in S1V2 (Figure 4a,b). The average CH₄ contents show that the pre-treatment successfully boosted performance in the whole stage. The average for S1 was 0.22 ± 0.01 dm³/gVS and 0.37 ± 0.02 cm³/gVS_removed, whereas S2 produced 0.25 ± 0.01 dm³/gVS and 0.40 ± 0.02 cm³/gVS_removed (Figure 4c).

Other researchers have also reported that SCO₂ pre-treatment significantly affects biogas production and CH₄ content. The difference was the most pronounced within the OLR range of 3.0–4.0 gVS/dm³·day and extended both to the output of gas metabolites and the CH₄ fraction. The study has also shown OLR to have a significant effect on these values. The highest daily yields from AD were recorded at 3.0 gVS/dm³·day (V2). S1 produced 17.7 ± 0.8 dm³ CH₄/day, compared to 20.4 ± 1.1 dm³CH ₄/day determined in S2.

In their review paper, Kazimierowicz and Dębowski [20] presented a summary of the studies carried out so far to assess the efficiency of biogas production from sewage sludge pre-treated with SCO₂. Based on their analysis, they identified areas requiring further research. The main one is the lack of research on anaerobic digestion in continuous reactors [20]. In literature, studies have only focused on the effect of SCO₂ pre-treatment of MSS on AD performed in batch reactors. As an example, Nowicka et al. (2014) [21] managed to boost biogas production by 49% using this technology [21]. Zawieja (2019) [22] employed the pre-treatment at a SCO₂ to MSS ratio of 0.55/1 (by volume), producing 0.62 dm³/gVS biogas containing approx. 78% CH₄ [22]. The non-pretreated MSS only yielded 0.43 dm³/gVS biogas. Evidently, biogas yield improvements were higher in batch experiments than in the present study. Many researchers have noted the relationship between the OLR and the yields of AD-produced biogas [5,30,31]. Elango et al. (2007) [30] examined the effect of different ORLs (0.5, 1.0, 2.3, 2.9, 3.5, and 4.3 gVS/dm³·day) on the AD of municipal waste mixed with municipal sewage. They achieved the maximum biogas production of 0.36 dm³/gVS·day at ORL = 2.9 gVS/dm³·day—being close to the optimal ORL level found in the present study. The methane content of the biogas ranged from 68 to 72% [30]. Researchers have pointed out that reactor overload can lead to diminished pH and the inhibition of methane-producing microbes, which are the main producers of biogas from acetic acid [32]. Hence, the OLR levels in excess of what can be absorbed by methanogens can lead to the cessation of methane production during AD, as well as to hydrogen and VFA accumulation [27]. Exemplary effects of the intensification of biogas/methane production from sewage sludge as a result of various pre-treatment methods are presented in

Table 2. Example effects of various methods of pre-treatment of sewage sludge before AD.

Type of Sludge	Pre-Treatment Method	Condition Pre-Treatment	AD Method	Effects	Ref.
Dairy activated sludge	Microwave (MW)	900 W, 12 min, 1814 kJ/dm3,	AD semi-continuous, 37 °C, HRT 15-day, 170-day	+57% biogas production	[33]
Thickened sludge	MW	2.45 GHz, 800 W, 1 min, 96 kJ/kg TS	AD semi-continuous, 37 °C, HRT 20-day, 67-day	+20% biogas production	[34]
Thickened sludge	Ultrasonication (US)	100 W, 8 min, 96 kJ/kg TS	AD semi-continuous, 37 °C, HRT 20-day, 67-day	+27% biogas production	[34]
MSS	US	3380 kJ/kgTS	Temperature phased anaerobic digestion TPAD-BMP assay, 55 °C, 37 °C	+42% methane production	[35]
Dewatered activated sludge	Thermal hydrolysis	65 °C, 6 bar, 20 min	AD pilot-scale, 37 °C, SRT 20-day	+30–40% biogas production	[36]
Secondary sludge	Thermal hydrolysis	134–140 °C, 3.4 bar, 30 min	AD batch, 35 °C, HRT 30-day	+40.2% methane production	[37]
MSS	Electrokinetic disintegration	34 kWh/m3	CSTRs, 37 ± 1 °C, SRT 20-day	+33% methane production	[38]
Concentrated sludge	High-pressure homogenization	150 bar	AD full scale, 36–38 °C	+30% biogas production	[39]
Secondary sludge	Fenton treatment	60 g H2O2/kg TS 0.07 g Fe2+/g H2O2	AD batch, 35 °C, 30-day	+15% methane production	[40]
MSS	Free nitrous acid-heat	nitrous acid 0.52–1.11 mg N/dm3, 70 °C	AD batch	+17–26% methane production	[41]
MSS	Acid treatment	8.75 cm3 HCl/kgMSS, pH = 2	AD semi-continuous, 35 °C, HRT 12-day	+14.3% methane production	[42]
MSS	Alkaline treatment	0.1 mol NaOH/dm3	AD batch (BMP), 21-day	+1.5% biogas production	[43]
Dairy sewage sludge	Pretreatment with SCO2	SCO2/sludge: 0.1–0.5, 4.5 ± 0.3–22.7 ± 1.3 Wh	AD respirometers, 42 °C, 25-day, OLR 5gVS/dm3, HRT 21-day	+20 ± 2.1–43 ± 3.2% biogas production	[14]
MSS	Pretreatment with SCO2	SCO2/sludge: 0.55/1	AD respirometers, 37 °C, 28-day	+44% biogas production	[22]
MSS	Pretreatment with SCO2 and alkaline	SCO2/sludge: 0.75/1 +2 M NaOH	AD respirometers, 35 °C, 21-day	+15% biogas production	[44]
MSS	Pretreatment with SCO2 and hydrodynamic cavitation	SCO2/sludge: 1/1	AD respirometers, 35 °C, 25-day	+62% biogas production	[45]

.

3.3. Organic Matter and Digestion Coefficient

Stage V1 and V2 produced the lowest levels of VS in the digestate. The levels for S1 were 64.6 ± 0.9% TS and 64.5 ± 0.8% TS, respectively. The S2 digestate contained 63.6 ± 0.6% TS (V1) and 63.9 ± 0.7% TS (V2) (Figure 5a). Increasing OLR to 5.0 gVS/dm³·day led to significantly higher rates of VS in the digestate, i.e., 70.2 ± 1.2% TS in S1 and 67.6 ± 1.4% TS in S2 (Figure 5a). There was a positive linear relationship between the OLR and VS removed. The biodegradation rate in S1V1 was 25.4 ± 0.2 gVS/day, which corresponded to 63.3% of the initial load (Figure 5b). Similarly, the mineralization rate in S2V1 was 67.1%, which translated to 26.8 ± 0.4 gVS removed per day (Figure 5b). At an OLR of 5.0 gVS/dm³·d (V4), the removal rates were 54.6 ± 1.2 gVS/day (43.4%) in S1 and 58.7 ± 1.3 gVS/day (47.9%) (Figure 5b). The VS mineralization rate was somewhat affected by the pre-treatment with SCO₂, but OLR was the main determinant.

The digestion coefficient/portion digested (η_F) decreased as the OLR increased in successive experimental variants (

Table 3. Digestion coefficient across experimental variants.

Stage	Digestion Coefficient (ηF) (%)
	V1	V2	V3	V4
S1	64.21 ± 1.2	64.57 ± 0.8	58.25 ± 1.0	44.88 ± 0.8
S2	67.75 ± 1.1	66.88 ± 1.2	65.85 ± 1.3	50.87 ± 0.9

). In S1, η_F peaked in V1 and V2 at 64.21 ± 1.2% and 64.57 ± 0.8%, respectively (Table 3). Subsequent variants of this stage showed a notable decline in η_F, with rates of 58.25 ± 1.0% in V3 and only 44.88 ± 0.8% in V4 (Table 3). On the other hand, S2 was characterized by higher η_F across all variants, ranging from 65.85 ± 1.3% in V3 to 67.75 ± 1.1% in V1 (Table 3). V4, however, marked a significant decrease in η_F, which fell to 50.87 ± 0.9% (Table 3).

The digestion coefficient—i.e., the organic removal rate and the final levels of organics in the digestate—is mainly shaped by the process parameters [46]. The stability of the digestate is primarily determined by OLR, HRT, and the process temperature [47]. Hydrolysis support and the pre-treatment methods also play a role [48]. A high digestion coefficient is desirable in order to reduce the environmental impact of the digestate (including putrescibility and odor emission) [49].

3.4. pH and FOS/TAC

The pH and FOS/TAC were found to correlate with the OLR. The pH showed a consistent negative correlation with organic matter levels and a positive one with HRT. Both stages showed a near-neutral pH (7.23 ± 0.04–6.96 ± 0.03) as long as OLR was maintained within 2.0–4.0 gVS/dm³·d (Figure 6a). An increase in OLR to 5.0 gVS/dm³·d triggered significant decreases in pH—6.75 ± 0.05 in S1 and 6.84 ± 0.03 in S2 (Figure 6a). Similarly, the highest tested OLR (V4) led to elevated FOS/TAC at 0.44 ± 0.02 (S1) and 0.48 ± 0.02 (S2) (Figure 6b). Such values are indicative of digester overload. The FOS/TAC ratio in variants V1–V3 ranged from 0.37 ± 0.02 to 0.43 ± 0.02 in S1 and from 0.33 ± 0.03 to 0.42 ± 0.01 in S2 (Figure 6b). In both stages, very strong positive correlations were found between the pH and biogas and methane production yields. The coefficients of determination reached R² = 0.9765 and R² = 0.9722 in S1 (Figure 7a,b) and R² = 0.9601 and R² = 0.9188 in S2 (Figure 7a,b). Similarly, very strong but negative correlations were found between FOS/TAC and biogas (R2 = 0.9287) and methane (R2 = 0.9365) yields in S1 (Figure 7c,d) and R2 = 0.9675 for biogas (Figure 7c) and R2 = 0.9316 for methane in S2 (Figure 7d).

The pH value is a key determinant of stability in an anaerobic reactor [50]. An almost neutral pH in an anaerobic medium means a balance between the hydrolysis, acidic, and methanogenic phases of anaerobic conversion [51]. The pH decreases are primarily driven by the build-up of VFAs, which are produced during the acidogenesis stage [52], which may occur with too high OLRs [53]. A decreased pH results in the inhibition of the activity of methanogenic bacteria, effectively reducing pollutant removal rates and biogas production [54]. Sanchez et al. [55] have also demonstrated that increased OLR—though a possible driver of methane yields—may lead to a decreased pH and system failure due to the accumulation of VFAs and free ammonia [55]. By the same token, the FOS/TAC ratio is used as a metric of AD stability. The indicator shows the ratio of volatile organic acids to the alkaline buffer capacity [56]. In a stable AD process, the FOS/TAC falls within 0.3–0.4 [57]. On the other hand, an FOS/TAC above 0.5 indicates a reduced production of biogas associated with the lack of optimal running parameters for anaerobic microbes [56].

3.5. Bacterial Community

The composition of the anaerobic bacterial community over time was not significantly affected by the SCO₂ pre-treatment. However, the abundance of methane-producing Archaea was indeed affected by the highest OLRs used in V3 and V4. With OLR at 2 to 3 gVS/dm³·day, the share of Archaea exceeded 24% in both stages (Figure 8). When the OLR range was 4 to 5 gVS/dm³·day, the proportion fell to 19–23% in S1 and 19–21% in S2 (Figure 8). Similar trends were found for Methanosarcinaceae (MSMX860). On the other hand, no relationship was found between the variant used and the abundance of Bacteria (EUB338) and Methanosaeta (MX825). The share of Bacteria in the community ranged from 68 ± 7% to 71 ± 8%, whereas that of Methanosaeta (MX825) accounted for 6 ± 2% to 8 ± 2% (Figure 8).

Similarly, feedstock pre-treatment prior to AD was found to have no effect on the anaerobic bacterial community in an experiment by Gagliano et al. (2015) [58]. As in the present study, the relative abundance of Archaea decreased as the OLR increased, regardless of whether raw or sonicated MSS was used. The application of Archaea-specific FISH probes showed the presence of long filamentous rods of the Methanosaeta spp. The latter was found in all digestion phases, highlighting the occurrence of acetotrophic methanogenesis, as expected in a mesophilic anaerobic system [58]. The present study results corroborate earlier findings reported by Boonapatcharoen et al. [59], who noted that an OLR increase from 1.0 to 6.0 kgCOD/m³·d during AD led to decreased abundance of Archaea [59]. Yet, another study [60] similarly reported that the anaerobic bacteria community was directly affected by rising OLRs, reduced HRT, and the resultant changes in reactor conditions.

3.6. Estimated Energy Production

The assessment of energetic effectiveness is extremely important to the evaluation of process viability, especially on the large scale [61]. Analyzing the amount of methane production and its energy value of 9.17 Wh/dm³, the highest gross energy gain was shown in S2V2 with an OLR of 3.0 gVS/dm³·day and amounted to 187.07 ± 1.5 Wh/day (

Table 4. Estimated energy production depending on the experimental variant.

			S1								S2
V	OLR	MMMS	Ymethane	Ymethane	CVmethane	Eout	V	OLR	MMMS	Ymethane	Ymethane	CVmethane	Eout	Eout(s2) − Eout(s1)
	gVS/dm3·Day	gVS/Day	dm3/kgVS	dm3/Day	Wh/dm3	Wh/Day		gVS/dm3·Day	gVS/Day	dm3/kgVS	dm3/Day	Wh/dm3	Wh/Day	Wh/Day
1	2	40	310 ± 14	12.4 ± 1.2	9.17	113.71 ± 1.2	1	2	40	340 ± 9	13.6 ± 1.4	9.17	124.71 ± 1.4	11 ± 0.2
2	3	60	270 ± 14	16.2 ± 1.3		148.55 ± 1.3	2	3	60	340 ± 19	20.4 ± 1.5		187.07 ± 1.5	38.52 ± 0.2
3	4	80	170 ± 11	13.6 ± 1.1		124.71 ± 1.1	3	4	80	220 ± 7	17.6 ± 1.6		161.39 ± 1.6	36.68 ± 0.5
4	5	100	110 ± 4	11 ± 1.1		100.87 ± 1.1	4	5	100	110 ± 10	11 ± 1.2		100.87 ± 1.2	0

M_MSS—mass of MSS; Y_methane—methane yield; CV_methane—methane calorific value; E_out—energy output; E_out(s1) − E_out(s2)—difference in energy output.

). In S1V2, only 148.55 ± 1.3 Wh/day was obtained. The use of pre-treatment with SCO₂ to intensify the methane fermentation process is justified when a closed CO₂ cycle is used in these solutions in the cycle: biogas production—biogas enrichment—SCO₂ production—sludge disintegration—fermentation—biogas production. This is an important argument that improves the economic, technological, and environmental efficiency of MSS anaerobic digestion [20].

4. Conclusions

The presented experiments are the first studies on the effect of pre-treatment with SCO₂ on the efficiency of anaerobic digestion of MSS in continuous reactors. The available literature data present only the results of tests carried out in batch reactors. So far, the impact of the applied OLR on the efficiency of MSS methane fermentation has not been assessed, which is also a novelty of the conducted research.

MSS pre-treatment with SCO₂ was found to be successful in the partial disintegration of microbial cells, as evidenced by the subsequent dissolution of organic matter and nutrients. The disintegrated MSS had higher levels of soluble COD, TOC, and mineralized nitrogen/phosphorus.

The present study shows that MSS pre-treatment with SCO₂ at an SCO₂/MSS ratio of 0.3 (by volume) significantly improves AD performance in terms of methane production and feedstock mineralization. The pre-treatment was found to have no negative effect on the long-term continuous operation of the reactor. The organic load rate played a much larger role in determining the ultimate performance of AD compared to the other variables tested.

The pre-treatment was found to enhance AD performance within the OLR range of 3.0–4.0 gVS/dm³·day. These values produced significantly higher biogas yields and CH₄ fraction contents, both in terms of daily production and per VS added. The most significant impact of SCO₂ on VS removal from MSS was found for OLR = 4.0 gVS/dm³·day. The SCO₂ pre-treatment was not found to affect pH, FOS/TAC, or the anaerobic bacterial community composition. Instead, these parameters were mainly affected by the OLR.

The highest gross energy gain of 187.07 ± 1.5 Wh/day was obtained at an OLR of 3.0 gVS/dm³·day. The production of SCO₂ in technologies dedicated to biogas upgrading and its use for MSS pre-treatment fits directly into the assumptions of the circular economy and corresponds to material recycling. It also supports the idea of reducing carbon dioxide emissions through its capture and use in a closed cycle.