Anaerobic co-digestion of landfill leachate with liquidised food waste using modified up-flow anaerobic sludge blanket reactor

Globally, growing concern over the management of landfill leachate (LFL) and food ZDVWH ): DULVHV IURPPHWKDQH &+ၹ DQGFDUERQGLR[LGH &2ၷ HPLVVLRQVZKLFK pose significant public health risks. However, LFL presents significant potential for biogas generation through the anaerobic co-digestion (ACoD) process. This technology provides significant advantages, enabling energy recovery while generating minimal residual sludge as a by-product. This study aimed to evaluate the co-digestion of LFL and liquidised food waste (LFW) under different mixing ratios and support carrier integration in an up-flow anaerobic sludge blanket reactor (UASB). A machine learning (ML) algorithm was employed to evaluate its performance for forecasting biogas production. The biomethane potential (BMP) tests 1, 2, and 3 were conducted for mono-digestion, co-digestion at various ratios, and different support carrier applications, using a total of 39 BMP test bottles. Both conventional and modified UASB reactors were operated at mesophilic temperature in a water bath with organic loading rate (2/5V) of 1 to 7 g/L/day in semi-continuous study phase 1, while LQSKDVHWKHPRGLILHGUHDFWRUVZHUHUXQZLWK2/5VRIWRJ/GD\7KHPRGLILHG 8$6% V\VWHPLQFRUSRUDWHVWKUHHW\SHV RI VXSSRUW FDUULHUVLQFOXGLQJ5HG&OD\%LR 5LQJ 5&%5 &HUDPLF%LR5LQJ &%5 DQG/DYD5RFN /5 7R IXUWKHUHYDOXDte biogas production, two activation methods, zinc chloride (ZnCl2) and sodium K\GUR[LGH 1D2+ ZHUH HPSOR\HG WR DFWLYDWH WKH VXSSRUW FDUULHUV 7R HYDOXDWH potential tools for forecasting biogas generation, artificial neural networks (ANN) and support vector machines (SVM) were developed based on BMP test data. The ANN model was developed using feedforward backpropagation, while a linear function was applied to the SVM model. The ACoD of LFL and LFW with a ratio of 65:35 has been selected due to its superior performance in biogas production during BMP Tests 1 and FRPSDUHGWRWKHUDWLRVRIDQG7KHSHUIRUPDQFHRIWKHPRGLILHG8$6% reactor was successfully boosted by incorporating a co-digestion of LFL and LFW, along with the addition of 1D2+-PRGLILHG&%5 VXSSRUWFDUULHUZKLFK UHVXOWHGLQ higher biogas production (3275 mL/d), specific methane production (SMP) (459.95 mL CH4J&2'added DQG FKHPLFDO R[\JHQ GHPDQG &2' UHPRYDO DW DQ 2/5RIJ/GD\GXULQJsemi-continuous study phase 2. The ANN model produced a FRUUHODWLRQFRHIILFLHQW 52 RIZKHUHDVWKH690PRGHOKDGDVOLJKWO\KLJKHU52 value of 0.930. This indicates that the SVM model had superior predicted accuracy in biogas production from the anaerobic digestion (AD) process, and it was also found to be a more reliable system to serve as a preliminary tool in commercial applications.

Text FK 2024 57 - Full Text.pdf Available under License Creative Commons Attribution Non-commercial No Derivatives. Download (7MB) Text FK 2024 57.pdf Restricted to Repository staff only Available under License Creative Commons Attribution Non-commercial No Derivatives. Download (8MB)

Actions (login required)

View Item