Lactic acid has been described previously in literature as a significant acid generated in the anaerobic digestion of ensiled energy crops such as maize silage or fermented vegetable wastes [7] [8] . Mrůzek and Groda [9] found that the methane content in biogas positively correlated with lactic acid contents. In dairy effluents, large quantities of lactic acid have been reported as a result of lactose degradation [10] [11] ; however the role of lactic acid in many biogas producing facilities is still largely ignored. It is also often absent in predictive mathematical models and software such as Aspen, Simba, Activated Sludge Model No. 1 (ASM1), Anaerobic Digestion Model No.1 (ADM1) etc. used for simulating biogas plants.

Biogas reactor failures have been reported previously in acidification reactors during transient overload conditions as a direct result of lactic acid accumulation [12] [13] . Due to the low acid dissociation constant of lactic acid [pK_a = 3.86], its accumulation can lead to dramatic drop in pH further resulting in breakdown of the whole digestion process [14] . Zhang et al. [15] recommended avoiding the accumulation of lactic acid to improve anaerobic digestion of kitchen waste-fueled biogas digesters.

Maize and its silage are the most ensiled crops in the world, widely used for biogas generation due to their high carbohydrate content and easy cultivability [16] . Fresh maize normally bears no lactate, however maize silage has been reported to contain lactic acid up to 10% dry mass [17] . These two carbohydrate-rich substrates were selected for this study due to their relevance to industry and difference in lactic acid concentrations. Additionally, lactic acid as a sole substrate was tested to assess its impact on biogas yield.

Anaerobic bacteria have a major role in the fermentation process, particularly in the hydrolysis of carbohydrates to monosaccharides. These monosaccharides are degraded to lactic acid, which is then converted to H₂ and acetate for further conversion to biogas [18] .

The structure and activity of the microbial community involved in fermentation often depends on the original inoculum source and starter biomass, as well as the operational and environmental conditions [19] [20] [21] [22] . Bacterial communities in biogas plants adapt to the substrates fed [23] . Hence, biological behavior becomes very specific to that particular plant, further influencing the biogas production [24] .

Three different inocula were selected for testing in this study; one from an agriculture-based biogas plant, one from a waste water treatment plant, and a final inoculum from a continually controlled and monitored standard reactor. The aim of this research was to investigate the role of starter inocula, lactic acid and substrate type on biogas yield, with a view to improving processes and predictive models of biogas systems.

Fresh maize and maize silage were sourced from the same farm in the Emden region of Lower Saxony, Germany. Fresh maize plant was freshly harvested and roughly shredded into 1 cm pieces and prepared maize silage was collected from the same farmer. All samples were frozen at −20˚C for later analysis.

Batch experiments were prepared in accordance with the German standard VDI 4630 [28] . In each 1100 mL bottle, 490 g of inoculum was mixed with 10 g of wet weight substrate. This inoculum ratio ensured an authentic biogas power plant feeding scenario [5] [20] . The two substrates had comparable dry mass content. Controls (reference) were prepared by mixing 10 g of distilled water with 490 g of inoculum. Biogas production was measured with an ANKOM’s (N1v0, 4RF2; RFS#194) gas production system with readings transmitted wirelessly to a computer every hour. Pressure values were further used to calculate the biogas production yield. This experimental setup was performed with the three different types of inocula, and experiments were prepared in triplicate.

The three inocula were checked for the biogas production with lactic acid as the sole substrate. D/L-lactic acid from Sigma Aldrich was used for the experiments to check for its degradation with the different sources of inoculum. Batch experiments were performed with 10 mM (0.9 g/L) of lactic acid using ANKOM wireless gas monitors as described previously. The average values for biogas production were presented. Lactic acid with its strong pKa value showed a remarkable impact by decreasing the pH of the inoculum. Therefore, the pH was adjusted to the original pH for each inoculum using NaOH. Gas chromatographic analyses were performed daily with samples taken from the side vent of the bottle.

Primary reactor feed used for the standard reactor was optimized to attain maximal biogas production in batch experiments. The energy source concentration required to achieve maximal biogas production was maize starch, rapeseed oil and casein protein with the concentrations 1.3 g/L, 2.0 g/L and 2.0 g/L respectively. Any concentration of energy sources above these proved to be inhibitory to biogas production. A lactic acid concentration of 10 mM neutralized D/L-lactic acid was used to demonstrate the impact of lactic acid as a sole substrate. These were made up to a final volume of 100 mL with standard inoculum for the comparative study. Experiments with 150 mM and 250 mM lactic acid were further performed in order to determine the inhibitory concentration. The reference for this experiment was 100 mL of standard inoculum. Batch experiments were performed for 3 hours using ANKOM wireless gas monitors. An activity curve was created in order to infer an approximate contribution of a small fraction of lactic acid to the total activity of the standard reactor with the optimized feed. The experiments were performed in triplicates, with readings determined every 5 minutes and the average presented.

Fresh maize and silage were considered for the experiments due to the major differences in their lactate content [29] . This difference in concentration was confirmed by the lactic acid analysis. Analysis revealed a negligible amount of D- and L-lactic acid in fresh maize, and 4.8% DM in maize silage (Table 2). Along

with this, minor amounts of acetic acid (accounting to nearly 12.5% of that of lactic acid) were detected in maize silage as well.

Inocula from agriculture-based biogas plant and the standard reactor showed the presence of L-lactic acid, suggesting these inocula may contain microbial populations adapted to survive in and utilise the substrate. Inoculum from the waste water treatment plant showed a negligible content of D- and L-lactic acid (Table 3). This could be explained by smallamounts of carbohydrates in the waste water source [30] .

A volume of 3.1L biogas was generated per day, with a total methane content of 68%. In the steady state, the metabolite showing the highest concentration was acetic acid (20 mM) which is a direct precursor of methane. Propionic acid (3 mM) and lactic acid (1 mM) were also detected. Starch can be enzymatically hydrolyzed to glucose and further fermented to lactic acid by lactic acid bacteria [31] . Therefore, presence of lactic acid in our standard reactor suggested starch fermentation and high consumption rate of the starch in the reactor [32] . Acetic acid was initially chosen to lower the pH to 4.0 to ensure sterility, however this led to its accumulation in the reactor and resultant pH drop issues. This problem was rectified byinstead selectingformic acid as an alternative. Sørenson’s buffer (0.133 M) was also found to be effective in maintaining pH stability in the reactor.

The 16S rRNA gene fragments of both bacteria and archaea revealed a clear difference in the microbial community composition of the starter inoculum from EWE biogas plant had BacteroidesandMethanosarcina, and the final standard reactor population, which was comprised of Firmicutesand Methanomethylovorans as predominant groups [33] .

In experiments using agriculture-based inoculum, maize silage showed enhanced biogas production compared with fresh maize (Figure 1(a)). This result correlates with previous studies of [34] [35] which report an enhanced biogas production rate with ensiled crops than fresh materials. There are several possible explanations for this. The hydrolyzed structure of silage can improve nutrient availability and the release of sugars to methanogens [29] . Sugars are pri-

marily fermented to lactic acid which are further transformed to acetate and hydrogen that can be used for methane formation [18] . Additionally, silage itself contains acids such as acetic and lactic acid which are easily degradable substrates for the microorganisms. Improved biogas production with maize silage has been reported previously as a result of heterofermentive lactic acid fermentation and presence of other organic acids [36] .

The inoculum from the waste water treatment plant failed to show any considerable difference in biogas production between the substrates (Figure 1(b)). The partial hydrolysis of maize and the presence of lactic acid were not the only precondition for enhanced biogas production. While in agriculture-based biogas plants the bacteria are more exposed to carbohydrate-rich substrates and are adapted to lactate, those in sewage waste water treatment plants might not be adapted to it, suggesting it is a non-adapted inoculum. Composition of starter inocula and the culture conditions have been reported to influence the methane production due to differences in the microbial composition [37] [38] . The role of the starter inoculum could thus be noticed to play a significant role in having an impact on the biogas production.

More biogas was produced from silage than fresh maize in the standard reactor experiments (Figure 1(c)). GC analysis further revealed an increased total acetic acid content during the degradation of silage compared to fresh maize (Appendix I). This might be due to the undergoing heterofermentative lactic acid fermentation [18] . Lactic acid was also found in the inoculum and this could indicate the presence of lactate producing, and probably also lactate- degrading microbial species in the reactor. The standard reactor was fed with maize starch as one of the major energy sources and this could explain its adaptability to maize and its degradation products.

The reference samples had also shown substantial biogas production in the absence of added fresh maize and maize silage. This could be due to the intermediate organic acid contents in those samples. Acetic, propionic, lactic, butyric and valeric acids were detected in all inocula (presented in Appendix II). Acetic acid was found to be dominant in each case, with concentrations in the range of 0.5 - 1.6 g/L across each of the inocula. The biogas produced during batch experiments with these reference samples was subtracted and is presented (Table 4) in order to provide an accurate contribution of biogas production with fresh maize and silage with each inoculum. This facilitated a clear comparison of the gas generated only from the substrates.

Factors such as pH, alkalinity, volatile fatty acids (VFA) have also been identified to influence the biogas production [2] . Inhibition at lower pH by free organic acids such as acetic, propionic, butyric and valeric acids have been reported by [39] . During our experiments, however, the addition of silage did not influence the pH remarkably. Inhibition by acidification was not observed in our batch reactor experiments. The acids were consumed fairly and a pH above 7 was observed for all the samples towards the end of the experimentation.

Maize and silage differ significantly in terms of lactic acid content as discussed previously. The silage is preserved against degradation by bacterial contamination as a result of the high concentration of lactic acid and the corresponding low pH, and also bears a more hydrolyzed structure [34] . There have been varying discrepancies in previous research surrounding the differences in biogas production with maize and its silage. With some reports demonstrating no difference in the biogas production with fresh and ensiled maize [40] [41] , the others report an enhanced biogas production with maize silage. Investigations by [29] and [30] report an increased methane yield by about 25% and 14.6% respectively with maize silage compared to green, fresh maize. Sträuber et al. [7] remark ensilage process serves as a pre-treatment measure for enhanced biogas production due to its hydrolyzed structure and acidogenic fermentation products. Works from [42] and [35] also show,enhanced methane production with lactate-rich silage. Such contradictions could possibly be due to the different nature and microbial composition of the inoculum. Numerous studies show that the microbial communities and their behavior in biogas reactors are controlled mainly by the starter inocula [20] [24] [37] [38] . Due to the importance of starter inoculum on the system, using silage as a primary substrate may have varying results in biogas production, as this is dependent whether the microbial community is adapted to lactic acid or not.

Biogas production from lactic acid as a sole substrate was measured to assess the impact of the acid alone on the system. Biogas was produced from lactic acid in cases where the microbial community present has the capacity to utilize the substrate. Maximum biogas production was observed by the inoculum from agriculture-based biogas plant (605 mL), followed by the inoculum from the standard reactor (460 mL) and the minimum biogas production observed in the inoculum from waste water treatment plant (210 mL) (Figure 2).

The inoculum from the waste water treatment plant displayed an inhibition of biogas production with no biogas production observed after an experimental run of 3 days. A reduced rate of lactic acid degradation was observed after 2 days, and GC analysis showed an accumulation of acetic and propionic acid. Propionic acid has been known to inhibit biogas production [43] which could explain the inhibition observed. The agriculture-based biogas plant inoculum showed almost complete utilization of lactic acid within the first day of the experimental run (shown in Appendix III) suggesting the presence of anaerobic bacteria with the ability to utilize the substrate. The standard reactor displayed a short lag phase compared to the other two inocula, with a reduced biogas production rate and lower total yield compared to the agriculture-based inoculum. This could be explained by a relatively low biomass, with a DM of only 1.8%, compared to 4.9% and 3% with the agriculture-based plant and waste water treatment plant inocula respectively. GC analysis revealed lactic acid reduction to 0 g/L after 3 days of experimental run.

Lactic acid was not a component of standard reactor feed; however the bacteria present in the standard reactor were able to convert starch present in the feed to lactic acid which was further utilized for biogas production. As lactic acid is an intermediate of sugar degradation [7] it is therefore reasonable to conclude that lactate is produced significantly during anaerobic digestion of carbohydrate- rich substrates.

Biogas production from the optimized feed for standard reactor was determined to compare it with that from a small quantity of lactic acid (10 mM) and further with larger quantities (150 mM, 250 mM). Optimized reactor feed showing maxi-

mal biogas production was found to be 3 times the concentration fed into the standard reactor. This energy source combination was comprised of starch (1.3 g/L) mixed with rapeseed oil (2.0 g/L) and casein protein (2.0 g/L). All other components of standard reactor feed remained unaltered.

Maximum biogas production was observed in the first hour for all treatments. Optimized feed produced nearly 48% more biogas than the reference treatment. It was found that the addition of just 10 mM of lactic acid had a large impact on overall yield and production rate, producing almost 35% of the biogas of the optimized feed. A concentration of 150 mM lactic acid resulted in the largest increase in total biogas yield. This exceeded the overall biogas production capacity of the reactor with optimized feed. Inhibition of biogas production was observed at 250 mM (Figure 3).

Obviously, the bacteria in this standard reactor are able to utilize lactic acid, even at high concentrations (150 mM). This is then rapidly converted rapidly to biogas. The considerable impact of lactic acid on biogas production and the overall system has also been demonstrated by the research of others, particularly in relation to silage-based anaerobic digesters [14] . With lactic acid resulting in such significant increases in biogas production, results of this study support our hypothesis that lactic acid indeed has a significant role in biogas production, and strongly influences biogas production rate and yield.