Ethanol may be produced from waste materials via a thermochemical-biochemical route employing gasification and syngas fermentation by acetogenic bacteria. This process is considered promising, but commercialization might be hindered by sub-optimal choices of design and operating conditions. In the present work, process systems engineering (PSE) techniques were applied for the optimization of a large-scale syngas fermentation bioreactor. Starting with the development of a dynamic model for a bubble column reactor with gas recycle, the multiple system outputs were studied with Principal Component Analysis to assist in the defmition of relevant objective functions, and artificial neural networks were used to approximate the steady-state responses with fast and accurate functions. This framework was then used to conduct a multi-objective optimization aiming at maximizing the ethanol production rate, lower heating value efficiency and ethanol titer, while also minimizing acetic acid titer and reactor volume.