Modelling of the biomethane production from foodwaste using anaerobic digestion

Abstract

Global urbanisation as well as economic growth has caused organic fraction of municipal solid waste (OFMSW) to be generated at an alarming rate. In developing countries this waste tends to be mishandled and disposed inappropriately at illegal dumping site and landfills. This lack of proper solid waste management negatively affects the environment due to the production of landfill gases and leachates. Food waste accounts for a larger portion OFMSW and its organic content is quite high. It is known to decompose easily when buried due its 80 % water content and this results in the creation of methane, a greenhouse gas. To counteract this, a process of anaerobic digestion can be used to treat foodwaste by generating biogas to be used for energy generation. This will also deal with the reduction of waste volumes at the dumping sites. The main purpose of this work was the biodegradability experiments, modelling and simulations of anaerobic digestion of food waste. This was achieved by two experimental. The first BMP tests were conducted on University’s canteen’s food waste which was used as substrate. The substrate had the volatile solids (%VS) of 21 %, moisture content of 78 % and total solids (%TS) of 22%. Two different types of inoculum, primary and digested sludge were used in batch reactors, at mesophilic temperature of 37 °C. Food waste inoculated with digested sludge resulted in methane yield of 357 Nml/g VS and its data fitted the first order kinetic model with k being 0.034 day-1, with R2 of 0.847. Whereas food waste inoculated with primary sludge also resulted in methane yield of 357 Nml/g VS and its data also fitted the first order kinetic model with k being 0.278 day-1 with R2 of 0.98. The second experiment fruit and vegetable waste was the substrate sourced from a local fruit and vegetable market. Pretreatment of the substrate (fruit and vegetable waste) preceded the BMP tests. The pretreatment method used in this instance was ultrasonic to assist in the improvement substrate’s biodegradability. Three sonication times of 10, 20, 45 min were under investigation; operated at an amplitude of 8 μm and 20 MHz .BMP tests were performed on ultrasonic pre-treated substrate which was inoculated with sludge. The pretreated samples contained volatile solids (%VS) of 82.1 %, total solids (%TS) of 17.9%, moisture content of 82.1 % and C/N of 26%. BMP tests in this instance were performed at mesophillic temperature of 35 °C. The sonication time of 45 min resulted in the highest cumulative methane production of 238 ±7.64 mL/g VS. This yield was 43 % higher than the untreated sample. 20 minutes for pretreatment sonication time, resulted in the cumulative methane production of 210.4 ±3.82 mL/g VS. Finally, sonication time of 10 minutes resulted in the least methane yield of 173 ±3.18 mL/g VS, with only 22 % increase in the yield. It was concluded that an increase in ultrasonic pretreatment time, resulted in an increased the methane yield. The last part of the study involved the use of a statistics software package, IBM SPSS Statistics, for curve fitting as well as an estimation of kinetic parameters of the models. The models which were investigated to fit the experimental data were original, modified and co-digestion Gompertz models as well as the original, modified and co-digestion Richards models. It was concluded that co-digestion Richard’s model fits the experimental data perfectly and it is recommended that this model can be used as is for fitting other substrates the have a two-step graph. modified Richard’s model is the perfect fit for all data under investigation. Codigestion modified Richards model has the highest correlation, with R2 of approximately 1 for all the fits, this is tabulated below. This shows clearly shows the perfection of the fit on the data at hand.