Biowaste Moisture as a Regulator of Carbon Monoxide Formation During Composting: Analytical and Microstructural Insights Toward Sustainable Waste Valorization

Abstract

Rising industrial demand for carbon monoxide (CO) motivates the development of sustainable pathways for its production. Composting has recently emerged as a potential biogenic CO source, yet the role of biowaste moisture in CO production has remained unquantified. In this study, the moisture dependence of CO generation during composting was assessed to address this knowledge gap. Laboratory-scale biowaste composting was conducted under mesophilic conditions (45 °C) with passive aeration for the initial 14-day phase, using three initial moisture levels: 31.6% (variant M100), 21.6% (M90), and 12.6% (M80), and periodic H₂O addition in M100 and M90. Monitoring of CO, CO₂, and O₂ concentrations, complemented by scanning electron microscopy of composts, revealed a non-monotonic moisture effect on CO formation. The intermediate-moisture treatment (M90; ~41–50%) was associated with the highest CO production, reaching a maximum of 681 ppm and 18.2 mg CO∙kg wet mass⁻¹, whereas high moisture (M100; ~51–64%) with lower CO levels (max. 276 ppm, 4.4 mg CO∙kg wet mass⁻¹), matrix compaction, elevated CO₂ and lower O₂ concentrations. The driest treatment produced trace CO (<20 ppm, max. 0.4 mg CO∙kg wet mass⁻¹) and retained a rigid, porous microstructure consistent with limited biodegradation. The results showed rapid but transient CO pulses after H₂O addition, implicating moisture-driven shifts in biological activity and/or abiotic formation. These findings identify an optimal moisture window for reproducible CO generation.

1. Introduction

In response to the increasing global demand for consumer goods such as plastics, textiles, electronics, automobiles, and renewable energy technologies [1], the need for industrial chemicals and metals continues to rise [2,3]. This trend directly fuels the demand for carbon monoxide (CO), which plays a vital role as a precursor in the synthesis of organic compounds and as a reducing agent in metallurgical processes [4]. CO is essential in the production of substances like methanol and acetic acid, as well as in the extraction of metals such as iron, copper, and lead [5,6,7,8]. The ongoing expansion of these industrial sectors underscores the increasing importance of CO in global manufacturing and resource processing. Reflecting this upward trajectory, the global CO market is expected to grow steadily, rising from an estimated value of $5.6 billion in 2022 to approximately $8.2 billion by 2032 [8].

This growing demand for consumer goods has simultaneously created the need to explore new, alternative, and sustainable methods of producing key chemical compounds, including CO, through the principles of the circular economy and bioeconomy [9]. These models promote the utilization of organic waste and residual biomass streams as valuable feedstocks for chemical production [10], thereby reducing reliance on fossil resources and enhancing resource efficiency in the chemical industry [11]. As a result, CO production through biological processes based on waste materials is becoming an increasingly attractive and strategic alternative for industry [12].

A recently emerging trend in the field of CO production from biowaste is the use of composting processes to generate CO [12]. Recent studies, building on earlier literature reports [13,14,15,16,17,18,19], have indicated that CO can be naturally released during the aerobic decomposition of organic matter in biowaste, offering a promising alternative to conventional industrial CO production methods, which are costly and challenging [5]. This biogenic pathway relies on specific bacterial strains that emit CO as a byproduct of their metabolic activity during composting [20]. However, the microbial community involved in composting is highly sensitive to fluctuations in environmental and process parameters, such as O₂ availability, temperature, organic matter content, pH, and moisture [21,22], making it a dynamic system for controlled CO production. Therefore, a thorough understanding of the optimal composting conditions for maximizing CO release is essential, and identifying and maintaining these parameters is crucial to ensure reproducibility, efficiency, and the potential scalability of this biogenic CO production pathway.

In recent years, the influence of temperature and O₂ availability on CO production during composting has been widely investigated. Studies have indicated elevated CO emissions under mesophilic conditions (45 °C) and thermophilic conditions (70 °C), as well as at 50 °C and 62 °C in the presence of Bacillus licheniformis, Bacillus paralicheniformis, and Geobacillus thermodenitrificans strains [20]. Regarding O₂ levels, research has shown that biogenic CO formation occurs under stoichiometric oxygen deficit, though its release is not a result of strictly anaerobic processes [23]. Furthermore, a recent study has explored the effect of organic matter content (OMC) in composted biowaste on CO production levels, demonstrating that even a relatively low organic matter concentration (dry OMC of ~20%) can lead to high CO concentrations (>2000 ppm) during composting [24].

Among the fundamental parameters influencing the composting process, the role of moisture level in shaping CO emissions has received no scientific attention. It is widely acknowledged that maintaining an optimal moisture level is essential for efficient composting, typically within the range of 40–60% [25]. Moisture strongly influences microbial activity; insufficient water content inhibits microbial metabolism, as well as the dissolution and transport of nutrients, and limits microbial mobility [26,27]. Conversely, excessive moisture restricts O₂ mass transfer, leading to anaerobic conditions due to reduced material porosity and an overly compacted matrix [28]. Despite the recognized importance of moisture level in composting processes, no studies have explicitly linked this parameter to CO production during aerobic organic matter decomposition.

To date, the only partially relevant insights stem from research on soils, media that share some characteristics with compost but differ in biological and physical properties. However, due to fundamental differences in structure, organic matter composition, and aeration conditions, these findings are not directly transferable to composting systems. Additionally, these studies are largely dated, with most originating from the 1970s, 1980s, and 1990s [29,30,31], and present inconsistent findings across different soil types. For example, Moxley and Smith [30] reported that optimal water content for CO production varied widely, ranging from 20% for deciduous woodland soil to 25% for coniferous woodland soil. Conversely, in Gullane soils and arable soils from Bush, CO emissions were negligible or even negative, with slight positive fluxes observed only at moisture levels of 15–20%. The authors also noted that dry soils generally acted as sources of CO, whereas wetter soils functioned as sinks. Interestingly, they observed a short-term increase in CO emissions following rewetting of previously dried soils—a phenomenon they initially attributed to moisture changes but later linked more strongly to temperature effects. Specifically, drying soils at 104 °C resulted in a 12-fold increase in CO emissions, compared to only a 2-fold increase at 25 °C, suggesting that high temperatures accelerate the thermal degradation of organic matter, leading to CO formation. More recent findings [32] corroborate this mechanism, showing that elevated temperatures can trigger spikes in CO production, further emphasizing the role of thermal processes in CO dynamics.

As for compost-specific literature, direct evidence remains scarce. The only notable mention appears in a speculative study by Hellebrand and Schade [13] from the early 2000s, where the authors proposed that moisture loss during composting may have contributed to a gradual decline in CO emissions over time. However, they also recognized that this effect might be confounded by the depletion of O₂ and reduced availability of molecular sources of CO in the organic plant matter. This further underscores the limited state of knowledge of CO dynamics connected with biowaste water content, specifically in composting systems.

Overall, despite the recognized importance of moisture in composting, its specific impact on CO emissions remains unstudied. This knowledge gap highlights the need for targeted investigations into how varying moisture levels influence CO fluxes during aerobic organic matter degradation. It was hypothesized that moisture content governs CO production during aerobic composting, with suboptimal levels constraining CO emissions through limitations in oxygen availability and moisture-induced changes in the structure and porosity of the compost matrix. Therefore, for the first time, this study aimed to examine CO production during laboratory-scale composting of a biowaste mixture, with particular emphasis on the initial mesophilic phase conducted at 45 °C under passive aeration, using experimental variants differing in the initial moisture level of the biowaste (32%, 22%, and 13%). Daily monitoring of process gas concentrations (CO, CO₂, and O₂) was complemented by scanning electron microscopy (SEM) analysis of compost microstructure to evaluate moisture-dependent changes in porosity and structural compaction, and to explore their implications for O₂ transfer and CO emission dynamics.

2. Materials and Methods

2.1. Biowaste Mixture

Three fractions of biowaste were used in the composting process: cherry branches (Prunus avium) obtained from spring tree pruning, cut into pieces smaller than 10 cm; grass clippings; and model food waste, prepared based on the percentage composition reported in the study by Valta et al. [33]. To obtain a C/N ratio close to the optimum for composting (25–30:1 [34]), these substrates were combined into a mixture in a fresh mass ratio of 4:1.5:1, respectively.

To achieve the target initial moisture levels of the biowaste for the three experimental variants: (1) natural starting moisture, (2) moisture reduced by 10 percentage points compared to the natural level, and (3) moisture reduced by 20 percentage points compared to the natural level, the prepared biowaste mixture was divided into three equal parts. The first part was proportionally distributed among four bioreactors numbered 1–4 (moisture variant ID: M100), representing the unmodified, natural moisture level (31.6%). The second part was dried in a laboratory dryer (KBC-65W, Wamed, Warsaw, Poland) at 105 °C for 30 min to obtain a moisture level of 21.6%, and then proportionally divided among bioreactors 5–8 (moisture variant ID: M90). The third part was similarly dried at 105 °C for 75 min, and subsequently distributed among bioreactors 9–12 when the moisture level obtained 12.6% (moisture variant ID: M80). The mass of biowaste in each bioreactor ranged from approximately 50 g in the variant with the highest moisture level to approximately 30 g in the variant with the lowest moisture level.

2.2. Laboratory-Scale Biowaste Composting

The laboratory-scale composting process was carried out in four replicates for each variant of initial moisture level in the biowaste (Figure 1), which is consistent with common practice in laboratory-scale composting studies, where inherent heterogeneity of the compost matrix leads to expected variability in gas emissions [25]. The experiments were conducted in airtight glass bioreactors, each with a volume of 900 mL, placed inside a thermostatic chamber (ST3, POL-EKO, Wodzisław Śląski, Poland) maintained at a constant temperature of 45 °C. During the experiment, the reactors were periodically repositioned within the chamber, which reduced potential positional effects and minimized systematic spatial bias [35]. Each bioreactor was sealed with a metal cap fitted with two hose connectors, to which high-temperature-resistant plastic tubing was attached. One of the tubes remained closed with a Hoffmann clamp throughout the process, while the other was equipped with a Mohr clamp and used for daily gas sampling. For this purpose, the tubing was connected to a gas analyzer (DP-28, Nanosens, Wysogotowo, Poland) to measure concentrations of CO, CO₂, and O₂. Due to the small volume of the bioreactors and a previous study confirming sufficient oxygenation under passive aeration in this type of bioreactor setup, a passive aeration strategy was adopted. This involved opening each reactor daily for approximately 15 min after measuring gas concentrations to allow for air exchange. To maintain the intended differences in biowaste moisture levels throughout the composting process, distilled water was periodically added after opening the bioreactors for aeration. In bioreactors 1–4, 10 mL of distilled H₂O was added every three days, evenly sprayed over the surface of the composting material. In bioreactors 5–8, the procedure involved adding 5 mL of distilled H₂O per reactor at the same interval. No additional homogenization was performed after H₂O application. The H₂O was sprayed uniformly over the entire surface of the composting material. Due to the presence of structural components (e.g., branches) in the mixture and the relatively small reactor volume, the compost matrix remained sufficiently porous to allow effective distribution of H₂O throughout the entire material without the need for mixing. No H₂O was added to bioreactors 9–12 throughout the process.

The composting process lasted 14 days, based on prior studies indicating that CO production tends to peak during the first two weeks and often declines sharply thereafter, eventually reaching negligible levels [12,23]. The mesophilic temperature regime (45 °C) was selected based on earlier findings showing that composting under these thermal conditions resulted in elevated CO levels [23,24]. Moreover, CO production has been linked to microbial activity in this temperature range, which is potentially influenced by the moisture level of the biowaste.

2.3. Process Gas Concentration Measurements

The analysis of process gas concentrations in the headspace of the bioreactors was conducted once daily at a fixed time over the 14-day composting period, before opening the bioreactors for passive aeration of the biowaste. Gas concentration measurements were performed using a DP-28 gas analyzer (Nanosens, Wysogotowo, Poland), which was connected to a plastic tube fitted to one of the ports on the bioreactor lid. The measurements included carbon monoxide (CO, ppm), carbon dioxide (CO₂, %), and oxygen (O₂, %). Each measurement lasted approximately one minute, allowing gas concentrations to stabilize. Once stable readings were achieved, values were manually recorded in a laboratory notebook. The analyzer was then disconnected to restore atmospheric baseline levels (CO ~0 ppm, CO₂ ~0%, O₂ ~20.2%). Measurement data were subsequently transferred to Excel for further processing and analysis.

2.4. Biowaste Sampling

Due to the relatively short duration of the composting experiment (14 days), sampling was conducted at three time points—on days 0, 7, and 14. This approach was selected as a compromise between data resolution and maintaining stable reactor conditions, since more frequent sampling could have disturbed the internal environment of the bioreactors, while less frequent sampling would have been insufficient to capture key process dynamics.

The initial sampling was performed on day 0, before the start of the experiment, to determine the physicochemical properties of the fresh biowaste mixture. These baseline measurements were intended to serve as a reference for tracking changes in gas composition, dry matter content, organic matter content, and other relevant parameters during the composting process, as included in Section 2.5.

During the composting process, on days 7 and 14, material samples were collected from each bioreactor for further analyses. For this purpose, following gas concentration measurements, each bioreactor was opened and, without unnecessary delay, a sample of the biowaste was taken using a sterile metal spatula and transferred into a sterile airtight container. The collected material was then subjected to the analyses described in Section 2.5 and Section 2.6.

Sampling on day 7 was carried out to capture the expected midpoint of the composting process, a period characterized by intensified microbial activity. This time point was chosen to monitor the progression of biodegradation and to enable the identification of early changes associated with different initial moisture levels applied to the biowaste mixtures. Final sampling on day 14 was performed to evaluate the overall progression of the composting process within the defined experimental timeframe. It was also intended to support the assessment of compost stabilization and to examine how varying moisture levels influenced the extent of material transformation under otherwise standardized conditions.

2.5. Substrates and Composts Characterization

Samples of the substrates and composts obtained after 7 and 14 days of the composting process, representing different moisture level variants, were characterized using standardized analytical methods. For samples collected on days 0, 7, and 14, dry matter content (d.m., %) was determined in accordance with PN-EN 14346:2011 [36], while dry organic matter content (o.d.m., % d.m.) was determined using Loss on Ignition (LOI) analysis, following PN-EN 15935:2022-01 [37]. In addition, elemental composition (C, H, N, S) was analyzed for the same materials using a Perkin Elmer 2400 Series analyzer (Waltham, MA, USA), in accordance with PN-EN ISO 16948:2015-07 [38]. For both the initial substrates and the final compost samples collected on day 14, respiratory activity (AT₄) was assessed using the methodology described by Binner et al. [39], employing the OxiTop Control system (Weilheim, Germany). Due to the substantial sample mass required for AT₄ determination vs. substrates mass in the bioreactors, intermediate sampling was not performed to avoid perturbation of reactor mass balance and composting process dynamics.

2.6. Scanning Electron Microscopy (SEM) Analysis

Before SEM analysis, the dried (105 °C, 24 h) compost samples were ground to a particle size of 70–300 mesh using a laboratory mill operating at 28,000 rpm (Chemland, Stargard, Poland). The samples were then mounted on the microscope stage using carbon tape. Their surfaces were subsequently coated with a gold (Au) layer by sputtering for 300 s (coating thickness up to 2 nm) using an Edwards Scancoat Six Sputter Coater (Manor Royal, Crawley, West Sussex, UK). Microstructural surface topography analysis was performed with an EVO LS15 scanning electron microscope (Carl Zeiss, Jena, Germany) at an accelerating voltage of EHT = 20 kV. Observations were carried out using a secondary electron detector (SE1). SEM micrographs were acquired for each sample at magnifications of 50×, 150×, 500×, and 1000×.

2.7. Analytical and Statistical Procedures

2.7.1. Event-Based Analysis of CO Concentration in Relation to H₂O Addition

To isolate and quantify the short-term effect of H₂O addition on CO concentration, independent of the overall temporal evolution of the composting process, an event study analysis was performed. The analysis included only reactors in which H₂O was periodically added (variants M100 and M90; Reactors 1–8). To account for potential differences between the two variants, the event study analysis was conducted separately for variant M100 (Reactors 1–4) and variant M90 (Reactors 5–8). Days of H₂O addition were identified individually for each reactor based on experimental records. Because H₂O was added at regular three-day intervals, the analysis window was restricted to a narrow time frame to avoid overlap between consecutive H₂O addition events. For each H₂O addition event, CO concentrations were extracted for a relative time window from two days before to two days after the event (−2, −1, 0, +1, +2 days), where day 0 corresponds to the day of H₂O addition. For each reactor and each H₂O addition event, CO concentrations were aligned to the event day (day 0). Data from all events were then pooled within each variant by relative day. For each relative day, the mean CO concentration and standard deviation were calculated across all available reactor–event combinations and used to construct the event study plots.

2.7.2. Calculation of CO Yield

The calculation of CO yield was carried out based on the methodology described by Sobieraj [24]. The CO concentration in the bioreactors’ headspace (ppm) was transformed into normalized mass values using Equation (1):

$${\mathrm{C}}_{\mathrm{g}\mathrm{a}\mathrm{s}}={\displaystyle \frac{{\mathrm{C}}_{\mathrm{p}\mathrm{p}\mathrm{m}}·\mathrm{M}\mathrm{W}·\mathrm{P}}{\mathrm{R}·{\mathrm{T}}_{\mathrm{r}}}}$$

(1)

where

${\mathrm{C}}_{\mathrm{g}\mathrm{a}\mathrm{s}}$—CO concentration, mg⋅m⁻³;

${\mathrm{C}}_{\mathrm{p}\mathrm{p}\mathrm{m}}$—CO concentration in parts per million, ppmv;

MW—molecular weight of CO, MW = 28 g⋅mol⁻¹;

P—atmospheric pressure, P = 101.32 kPa;

R—ideal gas law constant R = 8.314 m³⋅Pa⋅K⁻¹⋅mol⁻¹;

${\mathrm{T}}_{\mathrm{r}}$—temperature in bioreactor K = 318 K, K.

The headspace volume within the bioreactor (defined as the space above the substrates) was determined based on the previously measured bulk density of the composted mixtures and calculated using the following equation:

$${\mathrm{V}}_{\mathrm{h}\mathrm{e}\mathrm{a}\mathrm{d}\mathrm{s}\mathrm{p}\mathrm{a}\mathrm{c}\mathrm{e}}={\mathrm{V}}_{\mathrm{b}\mathrm{i}\mathrm{o}\mathrm{r}\mathrm{e}\mathrm{a}\mathrm{c}\mathrm{t}\mathrm{o}\mathrm{r}}-{\displaystyle \frac{{\mathrm{m}}_{\mathrm{b}\mathrm{i}\mathrm{o}\mathrm{w}\mathrm{a}\mathrm{s}\mathrm{t}\mathrm{e} \mathrm{m}\mathrm{i}\mathrm{x}}}{{\mathsf{\rho }}_{\mathrm{b}\mathrm{i}\mathrm{o}\mathrm{w}\mathrm{a}\mathrm{s}\mathrm{t}\mathrm{e} \mathrm{m}\mathrm{i}\mathrm{x}}}}$$

(2)

where

${\mathrm{V}}_{\mathrm{h}\mathrm{e}\mathrm{a}\mathrm{d}\mathrm{s}\mathrm{p}\mathrm{a}\mathrm{c}\mathrm{e}}$—headspace volume in the bioreactor, m³;

${\mathrm{V}}_{\mathrm{b}\mathrm{i}\mathrm{o}\mathrm{r}\mathrm{e}\mathrm{a}\mathrm{c}\mathrm{t}\mathrm{o}\mathrm{r}}$—the volume of the bioreactor, ${\mathrm{V}}_{\mathrm{b}\mathrm{i}\mathrm{o}\mathrm{r}\mathrm{e}\mathrm{a}\mathrm{c}\mathrm{t}\mathrm{o}\mathrm{r}}$ = 0.0009 m³;

${\mathrm{m}}_{\mathrm{b}\mathrm{i}\mathrm{o}\mathrm{w}\mathrm{a}\mathrm{s}\mathrm{t}\mathrm{e} \mathrm{m}\mathrm{i}\mathrm{x}}$—mass of the biowaste mix in the bioreactor, kg;

${\mathsf{\rho }}_{\mathrm{b}\mathrm{i}\mathrm{o}\mathrm{w}\mathrm{a}\mathrm{s}\mathrm{t}\mathrm{e} \mathrm{m}\mathrm{i}\mathrm{x}}$—bulk density of substrate, ${\mathsf{\rho }}_{\mathrm{b}\mathrm{i}\mathrm{o}\mathrm{w}\mathrm{a}\mathrm{s}\mathrm{t}\mathrm{e} \mathrm{m}\mathrm{i}\mathrm{x} \mathrm{M}100}$ = 265 kg∙m⁻³, ${\mathsf{\rho }}_{\mathrm{b}\mathrm{i}\mathrm{o}\mathrm{w}\mathrm{a}\mathrm{s}\mathrm{t}\mathrm{e} \mathrm{m}\mathrm{i}\mathrm{x} \mathrm{M}90}$ = 250 kg∙m⁻³, ${\mathsf{\rho }}_{\mathrm{b}\mathrm{i}\mathrm{o}\mathrm{w}\mathrm{a}\mathrm{s}\mathrm{t}\mathrm{e} \mathrm{m}\mathrm{i}\mathrm{x} \mathrm{M}80}$ = 243 kg∙m⁻³.

The daily mass of emitted CO in the bioreactor was determined using the following formula:

$${\mathrm{m}}_{\mathrm{C}\mathrm{O}}={\mathrm{C}}_{\mathrm{g}\mathrm{a}\mathrm{s}}·{\mathrm{V}}_{\mathrm{h}\mathrm{e}\mathrm{a}\mathrm{d}\mathrm{s}\mathrm{p}\mathrm{a}\mathrm{c}\mathrm{e}}$$

(3)

where

${\mathrm{m}}_{\mathrm{C}\mathrm{O}}$—the mass of daily emitted CO in the bioreactor, mg.

The results were normalized and expressed as CO yield (mg∙kg⁻¹ wet substrate). The calculations are presented in an Excel spreadsheet in Supplementary Material S1.

2.7.3. Data Analysis and Statistics

Statistical analyses were performed to evaluate the effects of moisture variant and time on gas concentrations (CO, CO₂, and O₂), as well as to compare peak CO values between variants.

Each reactor was treated as an independent experimental replicate within a given moisture variant. As repeated measurements were collected over time from the same reactors, bioreactor identity was included as a random effect in the statistical model to account for within-reactor dependence.

Time-series data were analyzed using linear mixed-effects models, with gas concentration (CO, CO₂, or O₂) as the dependent variable, and variant (M100, M90, M80), time (day), and their interaction as fixed effects. Separate models were fitted for each gas parameter (CO, CO₂, and O₂).

To support the evaluation of optimal moisture conditions, peak CO concentration was defined as the maximum CO value observed for each reactor during the experimental period. Differences in peak CO between variants were assessed using one-way analysis of variance (ANOVA). As assumptions of normality and homogeneity of variance were not consistently met, the non-parametric Kruskal–Wallis test was additionally applied. When significant differences were detected, post hoc pairwise comparisons were performed using Mann–Whitney U tests with Bonferroni correction to control for multiple testing.

Statistical significance was accepted at p < 0.05. All analyses were performed using Python (version 3.11).

3. Results

3.1. Biowaste and Composts Properties

During the 14-day composting process, the moisture level gradually increased across all studied variants (Figure 2). The adopted method of adding 10 mL of distilled H₂O every three days to variant M100 and 5 mL to variant M90 proved effective in maintaining proportional differences in moisture levels between these variants. In the variant with the highest moisture level, the difference between day 0 and day 7 averaged approximately 19 percentage points, further increasing by an additional 13.6 percentage points by day 14. For variant M80, the initial moisture level of 21.5% increased to 40.9% during the first week of the process and reached 50.4% on day 14. In the case of the M80 variant, the lack of distilled H₂O addition, due to faster natural drying of the compost, resulted in a greater difference between this variant and the M100 and M90 variants, and the moisture level itself did not show significant changes between days 7 and 14 (approx. 27%).

Organic dry matter content was high in all analyzed variants, exceeding 90% d.m. in the majority of samples (

Table 1. Substrates and composts properties (average ± standard deviation); nd.—no data.

Sample ID	Day of the Process	Organic Dry Matter Content, % d.m.	AT4, mg O2∙(g d.m.)−1
M100	0	85.01 ± 0.18	10.73 ± 0.88
M90		93.26 ± 0.41	9.41 ± 0.66
M80		92.44 ± 0.32	9.34 ± 0.57
M100	7	94.93 ± 0.98	nd.
M90		95.46 ± 0.61
M80		96.03 ± 0.66
M100	14	95.16 ± 0.74	25.66 ± 2.43
M90		94.00 ± 2.26	44.68 ± 0.20
M80		94.24 ± 0.69	20.63 ± 8.42

). The 14-day incubation period did not result in a substantial decrease in this parameter; the final values for variants M100, M90, and M80 were 95.16, 94.00, and 94.24% d.m., respectively. In contrast, a noteworthy effect was observed in the respiratory activity of the analyzed samples. For the substrates (initial samples, day 0), AT₄ values were low and exceeded 10 mg O₂∙(g d.m.)⁻¹ only in variant M100. In contrast, for M90 and M80, the indicator suggested low reactivity of the biodegradable material (AT₄ ≈ 9 O₂∙(g d.m.)⁻¹). After 14 days of the composting process, AT₄ values increased in all variants, reaching levels characteristic of unstabilized materials (approximately 26, 45, and 21 mg O₂∙(g d.m.)⁻¹ for M100, M90, and M80, respectively). This trend indicates the role of compost moisture in activating decomposition processes.

The elemental composition of the analyzed substrate and compost samples did not exhibit distinct or consistently directed trends (

Table 2. Elemental composition of substrates and composts (average ± standard deviation).

Sample ID	Day of the Process	C, %	H, %	N, %	S, %
M100	0	37.47 ± 2.61	5.19 ± 0.18	2.83 ± 0.44	0.76 ± 0.09
M90		39.71 ± 2.72	5.43 ± 1.15	2.25 ± 0.25	1.15 ± 0.23
M80		42.04 ± 1.12	6.02 ± 0.10	2.36 ± 0.23	0.81 ± 0.11
M100	7	45.79 ± 1.99	5.36 ± 0.77	2.02 ± 0.32	0.63 ± 0.15
M90		46.19 ± 2.84	5.72 ± 0.36	2.21 ± 0.59	0.68 ± 0.08
M80		46.07 ± 1.42	5.27 ± 0.63	1.98 ± 0.29	0.66 ± 0.27
M100	14	43.88 ± 5.35	4.87 ± 1.18	2.06 ± 0.34	0.64 ± 0.09
M90		47.53 ± 2.41	5.88 ± 1.08	2.42 ± 0.63	0.82 ± 0.17
M80		46.11 ± 2.30	6.30 ± 0.50	3.07 ± 0.48	0.93 ± 0.12

). Across all variants, the lowest carbon content was observed at the initial stage of composting (day 0), ranging on average from 37.47% in M100 to 42.04% in M80. After 14 days of the process, carbon content stabilized at approximately 44–48% on average (Table 2). No pronounced variations were observed in hydrogen content between days 0 and 14 or among the different moisture level variants. Hydrogen concentrations ranged from 4.87% for M100 on day 14 to 6.30% for M80 on the same day. A similar pattern was observed for nitrogen and sulfur contents. Nitrogen accounted for approximately 2–3% of both substrates and composts, while sulfur content did not exceed 1% in any sample, except for variant M90 on day 0 (1.15%). The raw data on the elemental composition of the analyzed biowaste and compost samples have also been published in an open data repository [40].

3.2. CO, CO₂, and O₂ Concentrations During Composting of Biowaste Mixtures with Controlled Moisture Levels

The results demonstrated that CO production during laboratory-scale composting of biowaste varied depending on the moisture level maintained in the bioreactor. The lowest CO release into the bioreactor headspace was observed in the variant with the lowest moisture level (M80). In this case, average CO concentrations during the process remained below 10 ppm, with maximum values not exceeding 20 ppm, and CO emissions were stable, showing no pronounced fluctuations between successive measurements (Figure 3). The most favorable variant in terms of CO concentrations was M90. Although the mean CO concentration over the 14-day process reached only 73 ppm, peak values increased up to 681 ppm and were observed predominantly during the second week of composting, for which the weekly average concentration reached 123 ppm. In contrast, variant M100 proved less favorable with respect to CO production. While peak CO concentrations ranged between 225 and 241 ppm, the 14-day average reached 47 ppm. CO generation in M100 was heterogeneous, with distinct peaks observed on days 4, 7, and 10 (276, 241, and 235 ppm, respectively), yet these values remained substantially lower than those recorded for M90.

A comparison of CO concentration profiles during the second week of the process between variants M90 and M100 indicated that moisture levels in the range of 41–50%, characteristic of M90, were more conducive to CO formation, whereas higher moisture levels of 50–64% (M100) exerted an inhibitory effect on CO release.

All tested variants exhibited high oxygenation levels, generally exceeding 17%, indicating an efficient compost aeration method in this case and the effective maintenance of aerobic conditions through passive aeration (Figure 4). Nonetheless, slight differences were observed among the variants. The M100 variant displayed the lowest O₂ concentration during the composting process, averaging 17.9% across the entire period. From day 7 onward, O₂ levels began to fluctuate, reaching a minimum of 15.1% on the final day. In contrast, the M80 variant was characterized by the highest oxygenation, with a 14-day average of 18.4% and peak values reaching 19.7%.

Although CO₂ concentrations remained low across all analyzed cases (averages not exceeding 4%), their levels and distribution within the bioreactors varied across experimental variants. In contrast to O₂ concentrations, the highest CO₂ levels were observed in the variant with the highest moisture level, where daily peaks exceeded 3% on day 8 and reached over 5% by day 14 of the process (Figure 5). The average CO₂ concentration over the entire period in this variant reached 2.7%. In the M90 variant, CO₂ concentrations dropped below 2% during the initial three days of composting, but in later stages they increased consistently, with peaks exceeding 3%. The lowest CO₂ concentrations were recorded in the M80 variant, where in some instances the gas was not detected at all.

The raw data on the concentrations of gases analyzed in this study can also be found in a publicly accessible open repository [41].

3.3. Short-Term Response of CO Concentration to H₂O Addition

To illustrate the response of CO production to distilled H₂O addition in variants M100 and M90, Figure 6 presents the mean CO concentration during the two days preceding H₂O addition (relative days −2 to 0) and the two days following H₂O addition to the composted biowaste (relative days 0 to +2), with the moment of H₂O application indicated on the x-axis as day 0. Interpretation beyond two days was excluded due to the regular three-day interval between consecutive H₂O additions.

The event study results indicate that following the addition of distilled H₂O to the composted biowaste, CO production responded rapidly, with changes already observable on the first day after H₂O addition in both moisture level variants. The baseline was defined as the mean CO concentration from days −2 and −1 to reduce the influence of short-term variability and better represent pre-intervention conditions. In M100, where a larger volume of H₂O was applied (10 mL vs. 5 mL), CO concentration increased on average by 31 ppm relative to the baseline (days −2 and −1,

Table 3. Effect size (ΔCO, ppm) following H₂O addition.

Variant	Baseline, ppm	ΔCO Day 0, ppm	ΔCO Day +1, ppm	ΔCO Day +2, ppm
M100	66.75	−49.81	31.06	−27.44
M90	74.25	7.81	20.81	−2.13

). In contrast, in variant M90, the increase was slightly lower, reaching an average of 21 ppm. However, the analysis further indicated that the stimulatory effect of H₂O addition on CO production was not sustained. On the second day after application, CO concentrations in both variants declined relative to both the baseline and day 0 values, with a more pronounced decrease observed in M100 (−27 ppm) compared to M90 (−2 ppm). However, it is worth noting that although variant M100 exhibited a larger relative increase in CO concentration after H₂O addition, variant M90 showed higher absolute maximum values and substantially greater variability, which was described in Section 3.2.

3.4. CO Yield

The analysis of CO mass in the bioreactors under different moisture content variants indicated the highest mean value for the M90 variant (0.06 mg), compared to 0.01 and 0.04 mg CO for M80 and M100, respectively (Supplementary Material S1). The maxima of CO production for M90 occurred in the second week of the process, reaching 0.57 and 0.47 mg CO (days 8 and 9, bioreactor no. 6; Figure 7, Supplementary Material S1). In contrast to M90, the M100 variant exhibited several peaks of CO production (on days 4, 7, and 10 of the process); however, these did not exceed the values observed for M90 (0.21, 0.19, and 0.18 mg CO, respectively).

The analysis of CO yield per kilogram of composted substrate showed that the maximum value reached 18.2 mg CO∙kg wet mass⁻¹ (M90 variant, day 10 of composting; Supplementary Material S1, Figure 8). For the variant with the highest moisture content, the maximum yield was more than four times lower (4.4 mg CO∙kg wet mass⁻¹, day 4 of composting), whereas for M80 it was 45.5 times lower (0.4 mg CO∙kg wet mass⁻¹, day 1 of the composting process).

3.5. CO, CO₂ and O₂ Dynamics and Statistical Analysis

The temporal dynamics of CO concentration were analyzed to assess the effects of moisture variant and time. A linear mixed-effects model, including reactor as a random effect, showed no significant main effect of variant (p > 0.05) or time (p = 0.935) on CO concentration. However, a significant interaction between time and the M90 variant was observed (p = 0.009), indicating that the temporal pattern of CO evolution differed significantly for this variant compared to the reference variant (M100). This suggests that while average CO levels were comparable across variants, the way CO changed over time depended on the moisture level, particularly in the M90 condition.

In addition to CO, the effects of variant and time on CO₂ and O₂ concentrations were evaluated using linear mixed-effects models. For CO₂, a significant effect of variant was observed. The M90 variant showed significantly lower CO₂ concentrations compared to M100 (p = 0.009), while a similar trend was observed for M80 (p = 0.080). No significant effect of time or interaction between variant and time was detected. For O₂, both M90 (p < 0.001) and M80 (p = 0.029) exhibited significantly higher O₂ concentrations compared to M100. No significant effect of time or interaction effects was observed.

To further evaluate the influence of moisture on extreme CO formation, peak CO concentration was calculated for each reactor. The analysis revealed significant differences between variants using the Kruskal–Wallis test (H = 7.79, p = 0.020), whereas one-way ANOVA did not detect significant differences (p = 0.122). This discrepancy is likely due to non-normal data distribution and high variability within groups. The M80 variant consistently exhibited low peak CO values (14–18 ppm), while M100 showed higher but relatively consistent peaks (133–276 ppm). In contrast, the M90 variant displayed the highest variability, including the maximum observed value of 681 ppm.

Post hoc pairwise comparisons did not reveal statistically significant differences between individual variants (p > 0.05). Nevertheless, a clear trend toward lower peak CO values in M80 compared to M100 and M90 was observed, although this did not reach statistical significance after correction for multiple comparisons. The lack of significant pairwise differences is likely attributable to the limited number of replicates and the high variability, particularly within the M90 group.

These results show that moisture level does not significantly affect mean CO concentration but has a pronounced effect on its temporal variability and the occurrence of extreme values. The M90 variant exhibited significantly different temporal behavior and the highest variability, suggesting less stable process conditions for CO production. In contrast, the M80 variant was associated with consistently low and stable CO levels.

Additionally, the relationship between CO and O₂ concentrations was examined. The assumptions required for Pearson correlation, including normality and linearity, were not met (Shapiro–Wilk test, p < 0.001), and visual inspection indicated a non-linear relationship. Therefore, no clear linear dependence between CO and O₂ could be established, suggesting that CO dynamics are governed by more complex, time-dependent processes rather than a direct relationship with oxygen concentration.

3.6. Surface Morphology of Biowaste and Compost Samples

The SEM analysis revealed pronounced moisture-dependent differences in the morphological evolution, structural integrity, and degradation pathways of the investigated biowaste materials. At the onset of composting (day 0), the highest-moisture treatment (M100) was dominated by well-preserved, elongated lignocellulosic fibers and coarse particles, with only a minor contribution of fine fractions (Figure S1), reflecting limited initial structural disruption. By day 7, substantial fiber fragmentation and a marked increase in fine particulate matter were evident (Figures S2 and S3), indicating intensified microbial and physicochemical degradation. Notably, particle edges became progressively rounded, accompanied by a reduction in surface roughness and a blurring of distinct structural boundaries (Figure S4), consistent with advanced lignocellulosic breakdown. Despite these transformations, macroporosity remained partially preserved, maintaining pathways for O₂ diffusion (Figure 9a). In contrast, by day 14, the M100 samples exhibited pronounced particle agglomeration and structural compaction. Fibers appeared flattened and partially collapsed, suggesting a significant loss of structural porosity (Figure 9b). The apparent filling of interparticle voids likely restricted O₂ transfer within the matrix. The surface morphology evolved toward a more amorphous, compact configuration, lacking clearly defined open channels and visible macropores (Figure 9c), indicating reduced aeration capacity under elevated moisture conditions.

Similar to M100, the initial biowaste sample of the M90 variant (day 0) was characterized by distinct, sharp-edged fibrous fragments (Figure 10a). SEM analysis revealed an open, porous structural organization of the substrate (Figure S5). Clearly visible macropores on the particle surfaces indicated a relatively low initial moisture level and a favorable aeration potential. By day 7, the onset of cell wall degradation became apparent. Despite progressing structural disruption, the compost matrix remained distinctly porous, with preserved plant cell lumina and no evidence of channel collapse (Figure 10b). The overall architecture retained its open configuration, suggesting sustained O₂ diffusion within the material. By day 14, M90 composts did not exhibit signs of structural pore closure observed in the higher-moisture variant. The compost structure appeared loose and spatially organized, locally sponge-like, yet still open with preserved pore networks. Macropores remained discernible in the SEM micrographs, confirming maintained structural permeability (Figure 10c). An increased proportion of fine particles indicated ongoing material decomposition (Figure S6). Moreover, pronounced degradation of cell walls and lignocellulosic fibers became evident, reflecting intensified biological activity under near-optimal moisture conditions.

The substrates and composts in the M80 variant exhibited a distinctly different morphological evolution compared to the corresponding materials in the M100 and M90 treatments. Although the initial substrate (day 0) displayed similar baseline characteristics—sharp particle edges, rigid fibrous fragments, and intact cell walls (Figure S7)—the subsequent composting process was dominated by physical disintegration rather than biologically driven structural transformation. Under the lowest moisture regime, no progressive particle aggregation, edge rounding, or cell wall swelling was observed. By day 7, the compost matrix was characterized by fissures, hollow tubular structures, and sharp fracture surfaces (Figure 11a), indicative of mechanical breakdown and dehydration rather than enzymatic degradation. Material desiccation was further evidenced by collapsed cell walls and the persistence of air-filled macropores without signs of pore filling or structural softening (Figure 11b). During the following week, no contour blurring or structural compaction was detected. Instead, SEM images revealed short fiber fragments, surface peeling, and layer delamination across the compost matrix (Figure 11c). Consequently, by day 14, the M80 compost retained a rigid lignocellulosic framework, reflecting limited biological decomposition under suboptimal moisture conditions.

All generated SEM micrographs of the analyzed substrate and compost samples are available in the open access data repository [42].

4. Discussion

Although the statistical analysis in this study showed that the moisture level did not significantly influence the mean CO concentration, it clearly affected its temporal variability and the occurrence of extreme values. In this context, the observed differences in CO release between the analyzed variants may reflect how moisture conditions shape the intensity and temporal pattern of CO production during composting. The analytical data and microstructural observations suggest that these responses may be associated with the dual nature of CO formation during composting—encompassing both biological and abiotic pathways—which may potentially depend on the moisture level of the composted material.

In general, the trend of increased CO production observed in the second week of composting for the M90 variant—and its non-decreasing pattern compared to the first week for M100—is rather unusual. In previous laboratory-scale studies on CO emissions during composting, conducted by Sobieraj et al. [23], CO production followed first-order kinetics. This meant that, after initially high values during the first 7 days, the CO emission rate gradually declined, reaching low levels during the second week and eventually approaching zero. It is important to note, however, that in those earlier experiments, once the process parameters (temperature and aeration) were established, the composting system was left undisturbed. In contrast to the present study, no interventions were made during the composting process. As a result, the CO production in the study by Sobieraj et al. [23] occurred naturally, governed solely by the initial process conditions. This may have led to a gradual depletion of the system’s capacity to generate CO. By comparison, the composting system of M90 and M100 variants described in this study was actively stimulated by the regular addition of H₂O, which contributed to the sustained or even elevated CO production in the second week of the process.

An important observation emerging from the comparative analysis of CO concentrations in the M100, M90, and M80 variants is that the highest CO yield was not recorded for the extreme moisture levels, but rather for the intermediate variant (M90). In the case of M80, characterized by the highest dry matter content, the biowaste was excessively dry from the onset of the process, potentially hindering biological decomposition. Over-drying during substrate preparation likely suppressed microbial activity or even induced dormancy, due to insufficient moisture to enable the transport of dissolved nutrients essential for microbial metabolism [43,44]. In both M80 and M90, low substrate moisture on day 0 resulted in physically stable biowaste (AT₄ < 9 mg O₂∙(g d.m.)⁻¹), yet biologically unstable (AT₄ > 20 mg O₂∙(g d.m.)⁻¹ by day 14). This aligns with the findings of Amuah et al. [45], who reported that such biological instability is typical when the moisture level drops below 30%.

Despite this common starting point, the dynamics of CO production diverged substantially between the two variants. In M80, CO levels remained consistently low throughout the composting process, which could have been attributed to the absence of H₂O supplementation. Although the moisture level naturally increased more than twofold by day 7—reaching 27% and remaining at that level until day 14—it still did not surpass the critical threshold of 30%. The trace-level CO production in M80 (in the range of several to a dozen ppm) was likely driven primarily by thermochemical, non-biological processes. This hypothesis is supported by findings from Stegenta-Dąbrowska et al. [46], who demonstrated that CO can be generated abiotically in sterilized compost at elevated temperatures. Similar to the current study, the abiotic CO emission observed by Stegenta-Dąbrowska et al. [46] exhibited a stable trend, with no major fluctuations. Furthermore, studies by Phillip et al. [17] and Stegenta et al. [19] confirmed that, in addition to heat, the presence of O₂ promotes thermochemical CO formation—an effect observed at technical composting scales. SEM analysis may further support this interpretation, showing that in this variant, no cell wall swelling, particle aggregation, or rounding of particle edges and contours occurred, and the rigid lignocellulosic structure of the biowaste was preserved throughout the 14-day process. Consequently, air-filled pores remained clearly visible, providing the highest oxygenation capacity among the analyzed treatments. Thus, the SEM micrographs indicated that physical disintegration predominated over biological decomposition. Despite the high O₂ availability, microbial growth and metabolic activity could have been limited due to suboptimal moisture levels, remaining below the threshold required for efficient microbial functioning. As a result, O₂ was not effectively utilized in biodegradation processes, contributing to the persistence of the highest O₂ concentrations observed during composting in the M80 variant.

In contrast to the M80 variant, which was potentially governed by abiotic CO production, the M90 and M100 treatments exhibited distinct, more dynamic patterns of CO release. In both M90 and M100 variants, CO generation did not follow a stable trend; instead, substantial fluctuations and multiple emission peaks were observed throughout the 14-day process. Such behavior may suggest a contribution of biologically driven processes. This interpretation is consistent with previous reports documenting bacterial CO formation during composting processes [20,46]. Based on the observations, it can be cautiously suggested that environmental conditions in the M100 variant may have been less favorable for sustained CO production compared to M90, although this cannot be conclusively demonstrated from the statistical results. During the second week of composting, moisture levels ranged from 50.6% to 64.2% in M100 and from 40.9% to 50.4% in M90. While these differences were not statistically linked to CO levels, they may have influenced process dynamics, particularly in terms of temporal variability and peak CO formation rather than average concentrations.

These observations may suggest that, in M100, bacteria responsible for lignocellulose and organic matter degradation—and potentially consequently CO release—did not experience optimal growth conditions due to inappropriate moisture level [47]. Elevated moisture levels in M100 likely limited O₂ transfer and diffusion within the compost matrix [48], resulting in the lowest O₂ concentrations recorded among all three analyzed variants. This interpretation is further supported by SEM analysis of M100 compost samples. SEM micrographs revealed reduced structural porosity and a compact matrix configuration restricting O₂ diffusion. The pronounced compaction and dense structural arrangement, particularly evident on day 14, may have promoted the formation of localized anaerobic or oxygen-limited microzones [49], as reflected by the highest CO₂ concentrations measured in the M100 variant during the final stage of composting.

In contrast, a more dynamic pattern of CO production was observed in the M90 variant, characterized by pronounced fluctuations and the highest recorded CO peak during the second week of composting. These observations may suggest that a moisture range of 40.9–50.4% could provide conditions favorable for processes associated with CO formation. However, this interpretation should be treated with caution, as statistical analysis did not confirm significant differences between variants in pairwise comparisons. This interpretation can be supported by SEM micrographs, which showed that despite the progression of the composting process, the compost matrix retained a well-developed porous structure. By day 14, no evidence of structural pore closure was observed, indicating the maintenance of adequate O₂ supply within the material [50]. Moreover, the moisture level in the M90 variant during the second week likely facilitated the transport of soluble compounds within the composting matrix and enhanced the physical and chemical accessibility of nutrients to microorganisms [51]. Together, these conditions could have promoted sustained microbial activity and intensified biologically mediated CO production.

Beyond the overall CO concentrations observed in the M90 and M100 variants, the stability of these potential CO-producing systems emerged as a critical factor. Analysis of the short-term response of CO concentration to H₂O addition revealed that a 5 mL dose in the M90 variant—compared with a 10 mL dose applied in M100—resulted in a more controlled and stable system behavior. In M90, the H₂O supplementation induced moderate fluctuations in CO levels, while maintaining overall system stability without pronounced emission spikes. In contrast, the M100 variant exhibited a rapid and transient short-term response characterized by sharp increases in CO concentration that dissipated quickly, typically on the second day. Such abrupt fluctuations suggest that stable environmental conditions conducive to sustained microbial growth and metabolic activity may not have been established. The pronounced and short-lived CO peaks in M100 may reflect a stress response of the microbial community to sudden moisture shifts, favoring survival-oriented metabolic adjustments rather than potential sustained CO-producing activity [52]. As a result of rewetting, active microbial populations may have been reestablished and could have progressively recolonized newly available microhabitats within the compost matrix, as previously demonstrated in studies investigating CO production in Scottish soils [30]. Conversely, the more moderate and prolonged response observed in M90 may suggest a balanced moisture regime that could support relatively stable microbial functioning and continuous biological CO production. However, this interpretation remains speculative, as microbial processes were not directly assessed in this study.

Additionally, it cannot be excluded that systematic rewetting of compost in the M90 and M100 variants may have stimulated not only biologically mediated CO production but also an additional abiotic component of CO generation. Similar conclusions were drawn by Conrad and Seiler [53], who reported findings consistent with those presented in the present study. In their work investigating CO fluxes from soils in subtropical regions, soil surfaces exhibited a relatively stable diurnal pattern of CO flux before irrigation, with concentrations remaining nearly constant. However, one day after rewetting, CO fluxes increased up to fourfold. Three days later, as most of the irrigation H₂O had evaporated, CO concentrations returned to baseline levels. The authors attributed this transient stimulation of CO production to chemical processes, hypothesizing that moist organic matter provides a more favorable substrate for abiotic CO formation than dry material. These findings may suggest that moisture-induced pulses observed in the present study may reflect a combined biological and physicochemical response, particularly under conditions of periodic H₂O addition.

The observed short-term response of CO concentration to H₂O addition also carries important practical implications. At the operational scale, particularly in full-scale composting facilities, irrigation of compost piles [54]—especially during the summer season and the initial, most intensive phase of composting—may trigger transient peaks in CO production. The pattern observed in this study indicates that H₂O addition triggers short-lived CO emission pulses rather than sustained increases. From a process management perspective, this suggests that irrigation strategy can be used not only for moisture control but also for controlling CO formation dynamics. In particular, the application of larger single H₂O doses and longer intervals between successive irrigations promotes the generation of higher and more pronounced CO emission peaks, enabling controlled increases in maximum CO concentrations and shaping their temporal profiles.

On the other hand, these findings indicate the need for enhanced process monitoring during periods of H₂O addition, as moisture-induced CO pulses may lead to temporarily elevated concentrations in the working environment. Given the toxic nature of CO and occupational health and safety considerations [55,56], careful control of irrigation practices and adequate ventilation strategies are essential to minimize potential risks to workers, particularly as previous studies have indicated the potential for exceeding occupational exposure limits in enclosed composting systems and similar setups [57,58,59,60,61,62,63].

The study has certain limitations. The experimental design does not allow for disentangling the relative contributions of initial moisture conditions and later-stage moisture convergence to the observed CO dynamics. Consequently, the extent to which CO differences are driven by early-stage conditions versus instantaneous moisture levels cannot be quantitatively resolved and warrants further investigation in future studies. Moreover, the initial oven-drying of the substrate at 105 °C may have altered its physicochemical properties [64,65], potentially confounding the attribution of observed effects solely to moisture differences.

Additionally, no direct microbiological analyses were performed, such as microbial community profiling, gene expression targeting CO-related metabolic pathways, or enzyme activity measurements. As a result, the presence and activity of microorganisms potentially responsible for biological CO production cannot be confirmed, and the proposed mechanisms remain hypothetical.

Another limitation is the relatively short duration of the experiment (14 days), which captures only the early, predominantly mesophilic phase of composting. While the results indicate that periodic H₂O addition may stimulate and potentially prolong CO emissions beyond typical early-stage patterns, it cannot be assumed that such effects would persist over longer composting periods (e.g., several weeks or months). Therefore, the findings should not be generalized to the entire composting process.

Furthermore, the experimental setup involved passive aeration, daily opening of reactors, and manual gas measurements, which may have influenced gas concentration readings. Reactor opening could lead to temporary dilution and disturbance of the gaseous phase, while manual measurements may have limited the temporal resolution and potentially missed short-lived fluctuations. Although all gas measurements were consistently performed before reactor opening, these factors may still introduce some uncertainty and should be considered when interpreting the results.

Finally, although the findings support the hypothesis that CO generation during composting could represent a potentially sustainable alternative to conventional, energy-intensive CO production pathways (e.g., waste gasification [66,67]), significant knowledge gaps and technological limitations remain. At present, there are no dedicated systems for controlled CO production from composting within biorefinery frameworks, nor established methods for managing microbial processes, capturing CO from gas mixtures, or its subsequent purification, storage, and distribution. These challenges must be addressed before such an approach can be considered practically viable.

5. Conclusions

This study shows, for the first time, that biowaste moisture may influence CO formation during composting and that CO yield may follow a non-monotonic response to H₂O addition. Under mesophilic conditions (45 °C) and passive aeration, the intermediate-moisture regime (~41–50%) was associated with the highest observed CO levels, reaching peak headspace concentrations up to 681 ppm and maximum yield of 18.2 mg CO∙kg wet mass⁻¹, whereas higher moisture (~51–64%) corresponded to lower CO (max. 276 ppm, 4.4 mg CO∙kg wet mass⁻¹), elevated CO₂ and reduced O₂ concentrations. The driest treatment (~27%) produced only trace CO (<20 ppm, max. 0.4 mg CO∙kg wet mass⁻¹) and showed limited biodegradation, suggesting that insufficient moisture may constrain CO formation and favor low, stable background generation. Furthermore, the H₂O addition triggers rapid but transient CO pulses, underscoring moisture as both a long-term regulator and a short-term perturbation driver. The SEM analysis provided mechanistic support for these findings. Excessive moisture level promoted matrix compaction and pore closure, plausibly restricting O₂ transfer, while near-optimal moisture preserved an open porous architecture that sustains microbial activity and CO production.

Future research should extend the composting duration beyond 14 days to capture late-stage CO dynamics and verify the persistence of the identified optimal moisture window. Microbial community profiling and targeted functional assays (e.g., metagenomics/qPCR of CO-related pathways, alongside sterilized controls) are recommended to directly apportion biotic versus abiotic contributions and to identify taxa and conditions associated with CO peaks. Finally, expanding compost quality assessment beyond AT₄—toward maturity/stability indices, humification metrics, nutrient and phytotoxicity tests, and contaminant screening—would strengthen the linkage between moisture control, CO formation, and the sustainability performance of compost-based waste valorization.