Improved Monitoring of Semi-Continuous Anaerobic Digestion of Sugarcane Waste: Effects of Increasing Organic Loading Rate on Methanogenic Community Dynamics

Abstract

The anaerobic digestion of filter cake and its co-digestion with bagasse, and the effect of gradual increase of the organic loading rate (OLR) from start-up to overload were investigated. Understanding the influence of environmental and technical parameters on the development of particular methanogenic pathway in the biogas process was an important aim for the prediction and prevention of process failure. The rapid accumulation of volatile organic acids at high OLR of 3.0 to 4.0 gvs·L⁻¹·day⁻¹ indicated strong process inhibition. Methanogenic community dynamics of the reactors was monitored by stable isotope composition of biogas and molecular biological analysis. A potential shift toward the aceticlastic methanogenesis was observed along with the OLR increase under stable reactor operating conditions. Reactor overloading and process failure were indicated by the tendency to return to a predominance of hydrogenotrophic methanogenesis with rising abundances of the orders Methanobacteriales and Methanomicrobiales and drop of the genus Methanosarcina abundance.

Keywords: biogas reactor overloading; methanogenic pathways; monitoring tool; stable isotope fingerprinting; sugarcane waste.

Figure 1

Methanogenic community dynamics in the mono- (a) and co-digestion (b) reactor. The relative T-RF abundance of methanogens in the digestate samples are given as function of experiment time. For each of the parallel reactors in the specific digestion set-up, two samples were analyzed, that in total four samples were analyzed for each, mono- and co-digestion. All samples belonging to the same digestion set-up had similar methanogenic community based on the relative T-RF abundances. Therefore, each bar on the graphic represents the T-RFLP profile calculated by the average of the four analyzed samples in each digestion set-up. The supporting clone libraries and sequence analysis of the selected clones allowed the taxonomic affiliation of the T-RFs from the community T-RFLP profiles of the complex reactor samples.

Figure 2

Isotopic dynamics of δ¹³C_CH4 (a); δ¹³C_CO2 (b); and δ²H_CH4 (c) along gradual OLR increase in biogas reactors fed with sugarcane waste products. Isotope data of CO₂ during the last overload phase contains data uncorrected regarding the pH shift induced degassing.

Figure 3

Characterization of the potential predominant methanogenic pathway along gradual OLR increase in biogas reactors fed with sugarcane waste products in mono- and co-digestion. In diagram (a) the dynamic shift of αC values are shown, while diagram (b) presents the correlation of δ²H_CH4 and δ¹³C_CH4; In (b) the dotted, dashed and lined hulls represent the beginning, middle, and end of the experiment, respectively. The numbers in the graphic indicate the experiment day.

Figure 4

NMDS analysis plot for correlating the T-RFLP profile of methanogens with the isotope composition of produced biogas. The smaller and bigger hull in the diagram represents the mono- and co-digestion in several sampling time, respectively. The letter M stands for mono-digestion and C for co-digestion set-up and the following numbers correspond to the sampling day. The dim grey and black arrows indicate the highly significant (p < 0.001) and significant (p < 0.05) correlations, respectively. Grey arrows indicate the correlation vectors of community differences and the isotope composition at lower significance (p < 0.5). Monte-Carlo permutation was used to test the significance against 999 random data sets. The direction of the arrows show the correspondence to the community structures and the length of the arrow indicate the strength of the correlation with the ordination axis.