Prediction models for flammability limits of syngas/air mixtures based on machine learning approach

Syngas is a promising hydrogen-containing energy source and industrial feedstock. However, because of its hydrogen content, syngas is prone to explosion if not handled properly. Moreover, the composition of syngas is complex and variable, and the explosion risk varies accordingly. In this study, the flammability limits of syngas/ air mixtures with added CH4, C2H6, C3H8, or NH3 were experimentally investigated. The flammability limit trends differed among these combustible gases, complicating prediction of the flammability limit. To enable the flammability limit of complex-composition syngas to be quickly and accurately obtained, the random forest (RF) and multilayer perceptron (MLP) models were trained on experimental data. Both models accurately predicted the flammability limits, but the RF was superior for the LFL prediction, and the MLP performed better for the UFL prediction. The proposed method for rapidly predicting the flammability limits of complex-composition syngas improves safety in the production and use of syngas.