Effect Of Adding Excess Sludge On Anaerobic Fermentation Of Fruit Residue

The citrus pomace waste generated is up to millions of tons each year in China, although it is belonging to the litter, there is a larger value in use. There has been related research on the anaerobic fermentation process of the citrus pomace home and abroad, and the research result shows that the process is feasible, which the methane produced can be used as energy and the biogas as fertilizer for farmland. However, the citrus pomace’s anaerobic fermentation alone can easily cause the acid accumulation and lowers the pH of the system, which affect the anaerobic microbial activity seriously, and resulting in the poor fermentation efficiency finally.The sludge is the same as the Marc, which belongs to the organic solid waste with the higher organic matter and moisture content. One hand, the slude has a strong buffering capacity because of the alkaline, On the other hand, it also is rich in N, P, K and the varieties of anaerobic microorganisms, which will improve the stability of the anaerobic fermentation system and the efficiency gas production after mixing with the citrus pomace, and reduces the cost in the dosage of alkali element and achieves the minimization of the solid waste.This experiment took the pomace waste which from a citrus fruit juice in Sichuan as our research target, using the semi-continuous fermentation and a regular daily access to materials, and sets the two reactors. The feed ratio which the Mixed group and the pure pomace group is 3:1, and provided the pure pomace feed group as the control one to compare and analyze the gas production and stability of the rector, and complementary researched the relevant indicators and influence level of the feed load on the anaerobic fermentation system, which determined the effective way to start the reactor and the best way of the feed load.Based on the principle of multiple regression, this paper used a large number of experimental data as the original data, and established the forecasting models of the gas production by the SPSS software for the correlation analysis, which to provide a reference for the application of practical engineering.The main conclusions are as follows through the experimental research:(1)The Marc-residual sludge mixed group was better than the Marc group in the removal efficiency of the COD, the ratio of the nutrients, the gas production and the system stability, which showed that the addition of the sludge contributed to the anaerobic fermentation of the citrus pomace.(2)The results showed that the best feeding load of the marc group and the Marc-sludge were 2.1kgCOD/m3.d and 2.5kgCOD/m3.d respectively, the minimum value of the CODcr concentration were 11012mg/L and 10045mg/L and the gas production 1409.7mL/d and 889.6mL/d, and the pH, VFA and ALK had a stable current. However, the gas production begun to decrease and there had no significantly improve significantly when the feed load of the two reactors were respectively more than 2.1kgCOD/m3.d and 2.5kgCOD/m3.d, the two systems started to lose stability.(3)The feeding load should adopt the incremental manner and start with the lower feed load feed, which contributed to the normal operation of the anaerobic reactor. The high load feeding mode directly was not conducive to the subsequent runs of the reactor, which easily lead to accumulation of VFA and the deterioration of the system stability.(4)The addition of the alkalinity could regulate the pH of the system timely and effectively. Yet, the new VFA couldn’t be transformed by the anaerobic microorganisms adequately with the feed continuing and the load rising, and resulting in the VFA accumulation, which affecting the fermentation process in the end.(5)This paper established two multivariate linear regression model in the base of the two sets of experimental data, and the R2 of the group 1was 0.997, the group 2 0.989, which proved that the goodness of fit were both good. The gas production predicted by the model’s had a accuracy rate of 82.3% and 81.5% respectively, which showed that the multiple linear regression model had some reference value.