Investigation of enhanced H2 production from municipal solid waste gasification via artificial neural network with data on tar compounds

An artificial neural network (ANN) is a biologically inspired computational technique that imitates the behavior and learning process of the human brain. In this study, ANN technique was applied to assess the gasification of municipal solid waste (MSW) with the aim of enhancing the H2 production. The experiments were conducted using a horizontal tube reactor under different parameters: temperatures, MSW loadings, residence times, and equivalence ratios. The input and output variables (released gases) were tested and trained using back -propagation algorithm, and the data distribution by K-fold contrivance. The values of the training (80% data) and validation (20% data) dataset were found satisfactory. The values of regression coefficient (R2) for the training phase were lied between 0.9392 and 0.9991, and 0.9363 and 0.993824 for the testing phase. Whereas; the values of root mean square error (RSME) for the training phase were lied between 0.4111 and 0.8422, and between 0.1476 and 0.7320 for the testing phase. Higher H2 production of 42.1 vol% was produced at the higher reaction temperature of 900 degrees C with LHV of 11.2 MJ/Nm3. According to the tar analysis, the dominant com-pounds were aromatics (17 compounds) followed by polycyclic aromatic, phenyl, aliphatic, aromatic hetero-cyclic, polycyclic, and aromatic ketone compounds.

Automated summary

An artificial neural network (ANN) was used to assess the gasification of municipal solid waste (MSW) and enhance H2 production. The experiments were conducted under different parameters and the input and output variables were tested and trained using the back-propagation algorithm. The results of the training and validation datasets were satisfactory.