Artificial Neural Network (ANN) Modeling for Hydrogen Production in a Continuous Anaerobic Sludge Blanket Filter (ASBF)

A laboratory scale anaerobic sludge blanket filter (ASBF) was operated for 217 days in two phases to investigate the effect of substrate concentration (2.53 – 13 g glucose/L) and hydraulic retention time (6 – 30 h HRT) at mesophilic temperature (32 ± 2 ̊C) and constant pH 5.5 for hydrogen production. At the optimized condition of 11 g glucose/L at 24 h HRT, 99.9% glucose degradation with 2.25 mol H2/mol glucose was achieved. The obtained data were modeled by an artificial neural network (ANN). The performance function determination coefficient (R = 0.9981) value obtained between experimental and simulated hydrogen production rate revealed that ANN model could reliably be used as a simulation model in ASBF.