Integrated process for simulation of gasification and chemical looping hydrogen production using Artificial Neural Network and machine learning validation

Abstract

A significant driver of global warming is fast growth in greenhouse gas (GHG) generated by energy-producing areas. Converting biomass into useful products, chemical looping gasification is an appropriate route. Using integrated steam gasification technology that utilizes a chemical looping method to produce hydrogen. Mn-, Ni-, and Ca-based materials are the three types of oxygen carriers (OC), and they are utilized. This process offers the syngas, in the greatest quality and quantity, which is a key factor. We consider that the optimum gasifier temperature is 1100 °C. The steam-to-biomass ratio is 0.95, if the steam is further increased then the char gasification reaction starts moving in the reverse direction. In this process, 736.629 MW of power is produced when only natural gas and air are used in the combustion chamber and if we add hydrogen, power is increased up to 16 MW. To predict syngas composition and the S/B ratio, machine learning modeling using Artificial Neural Networks (ANN) algorithms are applied and compared, Bayesian Regularization and Scaled Conjugate gradient proves to be the best ANN model for validating and comparing with process model, demonstrating its accuracy and potential for optimizing biomass gasification processes as high as up-to 0.99 R2 value.