In-ground lagoon anaerobic digester in the treatment of palm oil mill effluent (POME): Effects of process parameters and optimisation analysis

•Various machine learning models have been applied for prediction and optimization.•Machine learning models could be used to prevent digester upsets.•Include mixed liquor volatile suspended solids concentration as input parameter.•Further research on industry application and algorithm combination is...

Full description

Saved in:
Bibliographic Details
Published inFuel (Guildford) Vol. 357; p. 129916
Main Authors Yusof, Mohd Amran Bin Mohd, Chan, Yi Jing, Chong, Daniel Jia Sheng, Chong, Chien Hwa
Format Journal Article
LanguageEnglish
Published Elsevier Ltd 01.02.2024
Subjects
Online AccessGet full text

Cover

Loading…
More Information
Summary:•Various machine learning models have been applied for prediction and optimization.•Machine learning models could be used to prevent digester upsets.•Include mixed liquor volatile suspended solids concentration as input parameter.•Further research on industry application and algorithm combination is essential. Most researchers analyse and optimise the anaerobic performance based on the data generated from pilot plants and laboratory scaled experiments, but not from the industrial scale biogas plant. Therefore, this study aims to evaluate, compare, and optimise the performances of an In-ground Lagoon Anaerobic Digester under mesophilic conditions in four different biogas plants (BGP A, B, C and D) located in Malaysia, for total chemical oxygen demand (COD) removal and total biogas produced. Organic loading rate (OLR), temperature, recirculation ratio (RR) and plant location are found to have significant effects on COD removal and biogas production. Sensitivity analysis shows that BGP A, B, and C show a similar trend of sensitivity, i.e., OLR > Temperature > RR. Conversely, the temperature has the highest sensitivity contributing to the Chemical Oxygen Demand (COD) removal for plant D. Despite having a similar anaerobic digester design for each biogas plant, the optimum operating parameters of OLR, temperature and recirculation ratio vary from plant to plant. BGP B can achieve the highest COD removal (1.21 X 106 kg CODin/month) as well as biogas production rate (4.12 X 105 Nm3/month), with a corresponding methane yield of 0.28 Nm3 CH4/kg CODremoved when the optimal operating parameters for OLR, temperature and RR are configured at 1.60 kg CODin/m3 day, 37.7 °C and 2.08 respectively. It could be concluded that moderate OLR (<1.6), moderate T (<44 °C), and moderate RR (<2.3) are required to achieve optimum COD removal and biogas production for each biogas plant.
ISSN:0016-2361
1873-7153
DOI:10.1016/j.fuel.2023.129916