2.1 Sample characterization
SM had a TS content of 1.0%, VS of 59.5% (dry basis), pH of 7.0, a COD of 17,861 mg of O2 L− 1, total nitrogen of 1,200 mg L− 1, carbohydrate concentration was less than 2.5% (m/m), 0.2% of lipids (m/m) and 1.2% of proteins (m/m).
In PL, the measured TS content was 71.3%, VS, 73.2% (dry basis). Carbohydrate concentration was 14.9% (m/m), lipids, 0.8% (m/m) and proteins, 20% (m/m). The pH, COD and nitrogen were not measured in PL due to sample characteristics.
TS contents measured in SM were similar to those reported by other authors [40, 41], while VS were lower than those measured in other studies [15, 16], due to the management in cleaning the floor of the farms. The pH is in the optimal range for biodigestion and corroborates similar measurements in pig slurry [22] and SM [42] as well as COD and total nitrogen, with a value very similar to that measured for similar samples [23, 41, 43].
In PL, the measured TS are slightly lower than those reported in studies with this type of samples [6, 7]. This result is influenced by the inert material used in the litter and the number of batches housed.
Table 1 presents the results of the characterization of liquid fractions of SM with co-digestion of PL obtained after co-digestion and phase separation.
Table 1
Results of physical-chemical characterization in liquid fractions of co-digestion
Parameter
|
SM + PL
10%
|
SM + PL
20%
|
SM + PL
30%
|
SM + PL
40%
|
Co-digestion
|
TS (%)
|
8.1 ± 0.4d
|
15.1 ± 0.1c
|
22.1 ± 1.4c
|
29.1 ± 1.4a
|
VS (%) dry basis
|
71.6 ± 0.7a
|
72.4 ± 0.5a
|
72.7 ± 0.3a
|
72.9 ± 0.9a
|
Liquid fraction
|
TS (%)
|
4.3 ± 0.0d
|
7.8 ± 0.1c
|
15.5 ± 0.1b
|
25.2 ± 0.0a
|
VS (%) dry basis
|
65.5 ± 0.1c
|
66.1 ± 0.1c
|
67.8 ± 0.3b
|
70.7 ± 0.3a
|
pH
|
6.9 ± 0b
|
6.8 ± 0b
|
6.9 ± 0a,b
|
6.8 ± 0a
|
COD (mg of O2 L-1)
|
49,304 ± 21.7d
|
78,444 ± 13.5c
|
139,917 ± 101.9b
|
263,617 ± 35.1a
|
Total nitrogen (mg of N L-1)
|
1,250 ± 0d
|
2,000 ± 0c
|
2,700 ± 0b
|
5,517 ± 1a
|
SM: swine manure; PL: poultry litter; TS: total solids; VS: volatile solids; FS: fixed solids; COD: chemical oxygen demand. Values presented as mean ± relative standard deviation. N = 3. Means on the same line, followed by different letters show statistical difference between them according to the Tuckey test at 5% probability.
After phase separation, it was possible to obtain a sample with up to 25% TS (liquid fraction SM + PL 40%). Although it was evident the possibility of achieving different levels of TS with substrate co-digestion, it is important to point out that this achieved content is high for conventional biodigester technologies used in biogas production, which perform well with TS rates of up to 3% (covered lagoon) or CSTR (continuous stirred tank reactor), with TS rates at 10–15% [18, 44].
As for the VS content, it was possible to observe in co-digestion that there was excellent use of the organic fraction present in PL, which is evidenced by the small difference in the percentage of VS of liquid fractions of co-digested samples (Table 1). It is also noteworthy that there was no significant difference between VS contents of liquid fractions of co-digestion of SM + PL 10% and SM + PL 20%.
The pH results remained stable and within the range considered optimal for good activity of anaerobic microorganisms in the biodigestion process, around 7.0 [21], and similar to that reported by other authors, between 6.5 and 7.5 in anaerobic digestion systems [11, 18, 45, 46].
As for COD, as well as TS, it showed increasing values with the addition of PL percentages. The correlation follows what was stated for VS content of the transfer of organic matter present in PL to the liquid fraction of co-digestion (Table 1).
The N parameter also showed increment characteristics, favoring the balance of the C:N ratio, as the addition of PL increased, as observed in another study [25].
2.2 BIOCHEMICAL METHANE POTENTIAL
Thirteen BMP tests were carried out, which were carried out in the liquid fractions of the co-digestion foreseen in the Doehlert matrix experimental planning. Biogas and methane, which refers to the energy fraction present in biogas, were considered as experimental design responses. Table 2 shows the results.
Table 2
Levels of variables and responses obtained from the experimental design
Test
|
Variables
|
Answers
|
ISR
|
% PL
|
Biogas
(NL. kgVS− 1)
|
Methane
(NL. kgVS− 1)
|
1
|
1:1.25
|
0
|
593 ± 1a
|
402 ± 6a
|
2
|
1:0.5
|
10
|
435 ± 5b,c
|
286 ± 6b
|
3
|
1:2
|
10
|
350 ± 36f
|
233 ± 41e,f
|
4
|
1:1.25
|
20
|
405 ± 5b, c,d
|
268 ± 2b,c
|
5
|
1:1.25
|
20
|
386 ± 4d, e,f
|
255 ± 5b,c,d,e
|
6
|
1:1.25
|
20
|
445 ± 5b
|
284 ± 5b
|
7
|
1:1.25
|
20
|
358 ± 8d,e,f
|
235 ± 5c,d,e,f
|
8
|
1:1.25
|
20
|
394 ± 6c,d,e,f
|
263 ± 8b,c,d
|
9
|
1:1.25
|
20
|
406 ± 8b, c,d
|
260 ± 5b,c,d
|
10
|
1:1.25
|
20
|
395 ± 12c,d,e,f
|
258 ± 8b,c,d,e
|
11
|
1:0.5
|
30
|
353 ± 36e,f
|
211 ± 11f
|
12
|
1:2
|
30
|
355 ± 16e,f
|
219 ± 4e,f
|
13
|
1:1.25
|
40
|
400 ± 1b,c,d,e
|
228 ± 4d,e,f
|
ISR: inoculum substrate ratio; PL: poultry litter.
Means in the same column, followed by different letters show statistical difference between them, according to the Tuckey test at 5% probability.
The quadratic model Doehlert matrix showed an adequate fit, with satisfactory results (significant regression and non-significant lack of fit). Using experimental design results could be considered adequate, since the percentage of explained variation was 90%, and the maximum percentage explainable of 94% for the biogas parameter and 91% of explained variation and 95% of maximum percentage explainable for the methane parameter.
Figure 1 shows the interaction between the different liquid fractions of co-digestion with ISR for methane production. It is observed that the most favorable zone of the surface is the red one, which indicates the highest methane production, in which are responses of liquid fractions with 10–30% of PL and ISR 1:2, i.e., both variables, % of PL and ISR, influenced on the result. The optimal point of co-digestion was obtained with ISR 1:2 and addition of 20% PL. Similar results were obtained when biogas was used as an experimental design response.
In co-digestion of SM with PL [47], although without phase separation and with lower concentrations, an increase in biogas production was also recorded due to the addition of PL, which reinforces the possibility of using this material in anaerobic digestion.
When observing the specific biogas production, i.e., per kilogram of VS added to the experiment (Fig. 2), the best result was obtained with the highest ISR, in which there is a greater volume of inoculum in relation to the amount of substrate, similar to that found in the research by Dechrugsa et al. [15].
However, even with a low ISR, 1:0.5, no delay or inhibition of biogas production was observed (Fig. 2). This may be associated with the fact that the fractions were previously separated, which resulted in the removal of the lignocellulosic fraction present in PL and favored the biodigestion process, since lignocellulosic materials require adequate pre-treatment for good use in anaerobic digestion [27, 48]. In other words, it is an indicator of the relevance of the phase separation procedure proposed in this work.
Liquid fractions of co-digestion in which there were the best responses for biogas and methane production were those that presented TS contents within the operational conditions for covered lagoon and CSTR model biodigesters. More precisely, liquid fractions of SM + PL 10% (with 4.3% ST), SM + PL 20% (with 7.8% ST) and SM + PL 30% (with 15.53% ST) co-digestions meet the requirements for using the aforementioned biodigesters [18, 40]. When opting for liquid fraction of SM + PL 40% co-digestion, which also presented good performance in biogas production, a technological barrier is faced, because the TS rate of 25.24% is above what is accepted as ideal for a CSTR model reactor, which is 10–15% [18, 44].
The pure SM sample showed the best biogas and specific methane yield, with 593 NL kgVS− 1 and 402 NL kgVS− 1, respectively. As this sample is the easiest to degrade with the lowest rate of TS, it was a result within the expected range and very similar to that already found in a similar study [42], which with ISR 1:1 obtained 554 NL kgVS− 1 of biogas and 382 NL kgVS− 1 of methane.
In studies with ISR variation [42], a delay in the start of biogas production was reported in samples with a lower amount of inoculum, which was not observed in this study. In Fig. 2, it is possible to identify the effect of using the inoculum, since on the first day of the experiment there was an increase in the kinetic curve.
Still in kinetic analysis (Fig. 2), it is evident that the liquid fraction of SM + PL 30% co-digestion with ISR 1: 0.5 took longer to reach biogas production potential, when compared to the other samples. This type of behavior, however, was not observed in the liquid fraction of co-digestion of SM + PL 10, also with ISR 1: 0.5. Possibly, in this case, because the TS rate was lower (4.3% - Table 1) and, therefore, even though a smaller amount of inoculum was used, there was no damage to biodigestion. A possible delay in the action of microorganisms, reported by other authors [28–30] and associated with the ammonia present in PL, was not observed in the tests carried out.
2.3 INOCULUM MICROBIAL ENRICHMENT
The SM used for tests with EI showed 3.3% TS and 67.6% SV on a dry basis. In PL, TS content was 86.2%, and VS (dry basis), 62.0%. The TS measured in the SI were 3.2%, and SV (dry basis), 63.3%, while in the EI TS were 3.0%, and SV (dry basis), 62.9%.
Table 3 presents the inoculum enrichment results. The results obtained in the control standards, with both inoculums, met the recovery criterion established by the methodology used, since in both samples biogas production was greater than 596 NL kgVS− 1.
Table 3
Characterization of BMP samples and results
Parameter
|
CC (SI)
|
CC (EI)
|
SM + PL 20% (SI)
|
SM + PL 20% (EI)
|
ST (%)
|
96.0 ± 0.01a
|
96.00 ± 0.01a
|
9.54 ± 0.11c
|
9.54 ± 0.11c
|
VS (%) dry basis
|
99.9 ± 0.0a
|
99.9 ± 0.0a
|
54.23 ± 0.42d
|
54.23 ± 0.42d
|
Biogas
(NL. kgVS− 1)
|
698 ± 19.5a
|
705 ± 34a
|
355 ± 11.6c
|
537 ± 15.9b
|
Methane
(NL. kgVS− 1)
|
330 ± 4.2b
|
387 ± 12a
|
232 ± 0.6c
|
353 ± 17.2b
|
ST: total solids; VS: volatile solids; FS: fixed solids; SM: swine manure; PL: poultry litter; SI: standard inoculum; EI: enriched inoculum; CC: crystalline cellulose.
Means on the same line, followed by different letters show statistical difference between them according to the Tuckey test at 5% probability.
The characterization results of SM and PL samples, presented in Table 3, differ from the results measured in the first step of the experiment (Table 2). The main difference is in TS content of SM of 3.3% in this step and 1% in the previous step. This difference may be associated with the batch phase of housed animals or the management condition on the day of sampling. The same situation is observed for PL of 71.3% and 86.1%, respectively. However, despite this observed difference, in both cases, the results are consistent with those measured by other authors [6, 7, 16, 49, 50].
As for the inoculum’s TS content, the data presented in Table 3 indicate that there was little difference between SI and EI. In other words, despite the addition of broth enriched with microorganisms, such volume was not significant in the TS rate.
In the sample of the liquid fraction of co-digestion (SM + PL 20%), there was a marked increase in biogas production, when using EI, of 34%, with 355 NL kgVS− 1 for the sample with SI and 537 NL kgVS− 1 for the sample with EI (Table 3). As for methane, when analyzing the percentage as a function of the total biogas produced, content was 65% in both cases. This result is similar to that observed in a study with sludge enrichment with animal rumen [32].
As for the bacteria isolated and used in inoculum enrichment, in research on the microbiota of reactors, some of them are reported by other authors such as Rhodococcus sp, in the conversion of lignocellulosic products into lipids, which are easier to degrade [51], and Vagococcus acidifermentans, in acid production from less complex carbon sources [52]. Lysinibacillus sp. and Proteiniclasticum sp. were identified in significant concentration in an anaerobic reactor for wastewater treatment from a brewery and in a reactor with different substrates, respectively [53, 54].
Figure 3 shows the kinetic profiles of samples and reference standard with both inoculums. Despite the higher biogas production, no relevant change was observed in degradation kinetics when comparing the samples with both inoculums. The peak of biogas production was reached at similar times, i.e., the gain from inoculum enrichment was, in fact, in the greater biogas production and not in the anaerobic digestion process acceleration.
There is a difference in the degradation profiles in the experiment’s initial step, with a slight delay in the beginning of biogas production from cellulose with both inoculums, but from the third day onwards this no longer occurs, and the degradation process proceeds normally. Thus, it is possible to conclude that, although the SI has not been characterized in terms of its microbiota, it is evident that microbial enrichment was satisfactory to favor consumption of more complex organic matter, since for microcrystalline cellulose there was no difference in the volume of biogas produced.
Although microbial enrichment has shown good results in biogas production in the tested sample, it is important to highlight that the anaerobic digestion process is symbiotic and harmonious. Thus, the evolution of research in the area of microbiology applied to anaerobic digestion is necessary for later application on a full scale.