3.1. AD Trials
Following the schematic origin of the different faeces samples (A and B), their physicochemical characteristics are reported in
Table 1.
Regarding the TS values, it is possible to state that sample B has a lower value than sample A (a decrease of around 5%). However, the nitrogen content of the B sample was slightly lower than that of the A sample, showing an improvement of 7% in the C/N ratio, and, consequently, can be beneficial for the fermentation process.
These findings are consistent with some authors’ [
35] reports of C/N ratio values for post-weaning piglet slurry in the range of 14.5, as well as with other authors’ [
36] usage of pig dung and presentation of a C/N ratio value of 13.45. To make samples A and B viable for bioconversion in a wet AD system, it was necessary to adopt some pre-treatment procedures. The objectives for the selection of the pre-treatment were to obtain substrates (S
0 and S
1) with similar physicochemical characteristics to pig slurry, mainly the OLR, to allow a comparison with previous studies [
37,
38].
Table 3 shows the physicochemical characteristics of the feeds and digestates obtained in AD trials.
The results in
Table 3 highlight the positive impact on the pH of the AD feed due to the incorporation of
Laminaria digitata. This increment is favourable for the bioconversion process, in accordance with the values recommended for AD (6.5–7.5) [
39].
The 15% rise in the C/N ratio that resulted from the addition of
Laminaria digitata to the feed is another important factor in improving the AD process. This value is closer to the range of values that are ideal for the AD process, which is between 15 and 30 [
40]. It is important to achieve a well-balanced C/N ratio since methane production will be low due to limited carbon and increasing ammonia concentration if the C/N ratio is unbalanced [
41]. Comparing C/N ratios from this experiment with a study carried out with pig slurry from the fattening/finishing phase [
38], AD
0 corresponded to an increase of 86%, and AD
1 doubled this parameter.
The TS and VS values are in line with those reported by previous authors [
42] (34.6 g/kg and 24.5 g/kg, respectively), who used pig manure as the main substrate. The same applies to the TCOD value, which, in the same study, was 33.7 g/L.
The statistical analysis of AD0 and AD1 characteristics confirmed that there were no significant differences (besides pH and EC) between trials (p > 0.05), meaning that faeces from conventional diets can be replaced by faeces from piglets fed diets fortified with 10% L. digitata.
Regarding the TCOD and VS removal efficiency, the performance of the two mono-digestion runs was similar, showing values of 67–68% and 68–71%, respectively. In both trials, the pH of the digestate indicated the AD process’ stability.
During AD trials, the definition of key operational parameters is crucial to evaluating the performance and stability throughout the AD
0 and AD
1 assays.
Table 4 summarises the values obtained to allow a comparison between AD
0 and AD
1.
As shown in
Table 4, the OLR values did not have significant differences during the two trials. For other operational parameters, such as the GPR and MPR, the values presented also do not have significant differences, so the process performance was not affected by the incorporation of 10%
Laminaria digitata into the feed, as supported by the statistical analysis (
p > 0.05). The SELR value is below 0.4 d
−1 for both trials, indicating that the ratio between the organic load and the reactor’s biomass is appropriate for a stable bioconversion process [
30].
3.2. AcoD Trials
Given the improvement of the feed characteristics due to the introduction of L. digitata, co-digestion trials were performed using feed S1 as a substrate.
The co-substrate was defined based on a previous work [
23], where it was concluded that NF concentrates from mango peel could be utilised as a co-substrate, so the characterisation of this concentrate was carried out to assess its incorporation in AcoD trials.
Table 2 presents the characteristics of S
2, which facilitate the process of selecting the proportion of the co-substrate suitable to incorporate with the selected substrate of AD.
In
Table 2, it can be observed that the organic matter present in co-substrate S
2 (nanofiltration concentrate) is mainly composed of monosaccharides and disaccharides, comparing their concentrations with that of volatile solids, VS, and with results obtained elsewhere [
23]. The concentrations of TKN were lower because most of the organic nitrogen was retained by the previous ultrafiltration process. Using a similar strategy to that utilised in this study, other researchers [
43] who attempted to recover polyphenols from spinach and orange by-products came to the same conclusion that the polyphenols could be separated from simple sugars using NF membranes with a cut-off of 300 Da. In this case, the polyphenols were retained by NF membranes, while simple sugars were recovered in permeates [
43]. In this research, the use of NF membranes with a much lower cut-off (130 Da) allowed the retention of the most fermentable sugars in the concentrates, as intended for the anaerobic co-digestion process. In fact, the C/N ratio obtained for co-substrate S
2 was much higher than that obtained for S
1 (
Table 3).
This factor guarantees that S
2 will be an adequate co-substrate for S1, as it will aid in increasing the C/N ratio of the feed mixture to a ratio that is between 15 and 30, which is more favourable for the anaerobic digestion process [
40].
The use of 20% integration of S2 was chosen to provide a balanced synergetic effect and to follow recommendations for the AcoD process, since the pH of S2 is lower than the process’ optimal value, and, in contrast, the high C/N ratio value (56) could endanger the feed mixture.
To support the above statements, a ratio of 80:20 S
1:S
2 is suitable to maintain the performance and stability of the AcoD assay. With this choice, the feed combination shown in
Table 5 displays pH and C/N ratio values that are within the range advised by other researchers [
39,
40].
Table 5 presents the feed mixture’s physicochemical characteristics, as well as those of the digestate, after the AcoD process.
The most relevant change resulting from the introduction of S
2 as a co-substrate was a 20% increase in C/N, as expected, because NF concentrates from mango peel mainly contributed carbon (as seen in
Table 2). S
2 resulted from a sequence of membrane processes (UF-NF), where most of the organic matter was retained by UF membranes. As a result, NF concentrates have a reduced solids content, which lowers feed VS by about 17%. The C/N ratio reached in AcoD experiments increased by 80% when compared to prior research employing pig slurry and pineapple peel waste, with an OLR of 1.45 ± 0.02 g VS/L
reactor.d [
38].
Table 6 summarises the performance and stability parameters for AcoD trials, which allowed the evaluation of the effect of the incorporation of the 20% NF as a co-substrate.
It is important to mention that the OLR of AcoD is 34% lower than that of AD
1 (2 g VS/L
reactor.d,
Table 4).
It was feasible to assess the process’ stability by comparing the SELR values between AD
1 (0.38 ± 0.03 d
−1) and AcoD; the biogas quality exhibited the same behaviour (expressed as a percentage of CH
4). These results confirm that the process was not compromised by the change to AcoD. To determine the effect of the incorporation of both
L. digitata and the co-substrate, a statistical analysis was performed on the pH values of the feed and digestate. To represent this analysis, a box-and-whisker plot was created (
Figure 4) with a five-number summary: the minimum, the 25th percentile, the median, the 75th percentile, and the maximum, with the whiskers extending to the minimum and maximum.
In
Figure 4i, it is possible to observe the positive effect of
L. digitata incorporation on the feed pH. The control trial (AD
0) had pH values of 5.4 ± 0.2, and, with macroalgae incorporation, these values had an increase of 20%, reaching 6.5 ± 0.3. Based on the statistical analysis performed, it is possible to conclude that there was a significant increase (
p < 0.05). This fact is relevant because pH values should be in an ideal range for anaerobic digestion to occur steadily and increase biogas generation (6.5–7.5). Below 6.5, the methanogenesis growth rate could be reduced, and for values higher than 7.5, a system failure can occur [
39].
For AcoD, although the co-substrate addition had a positive effect on pH values (6.7 ± 0.3), this difference was not statistically significant (p > 0.05).
Regarding the pH values of the digestate from AD trials (
Figure 4ii), it is possible to observe that there were no major fluctuations: the pH in the trials remained near neutrality [
38]. However, when statistically analysing the values, in contrast to the feed pH, there were no significant differences between AD
0 and AD
1 (
p > 0.05).
As mentioned before, the SMP parameter was suitable to compare the performance of the different trials to offset the variation in the OLR.
Figure 5 illustrates the average SMP values for each trial.
During the AD
1 trial, an increase of about 8% in methane yield occurred in comparison with the AD
0 assay. Values from other studies are in line with those achieved in this experiment: one of them reports an SMP of 220 mL/g VS for AD trials performed with pig faeces with a control diet [
41]; another one, also for AD with pig slurry, indicates an SMP of 212 mL/g VS [
39].
The adoption of a co-digestion regime led to a more relevant increase in SMP of 29%, indicating that bioconversion was more efficient. This fact is probably associated with the more balanced C/N ratio caused by the co-substrate addition. This statement is confirmed by other authors who noted that combining co-substrates with animal manure improved the C/N ratio, which led to an increase in methane yields [
8,
39]. For example, the co-digestion of pig slurry with pear waste led to an SMP of 243 mL/gVS, which corresponded to an increase of 35% compared to pig slurry mono-digestion [
4].