Simulation of Supercritical Water Gasification of Organic Wastes with Neural Network Model

  • Qinming Zhang, State Key Laboratory of Multiphase Flow in Power Engineering, Xi'an Jiaotong University, China, China
  • Prof Shuzhong Wang, State Key Laboratory of Multiphase Flow in Power Engineering, Xi'an Jiaotong University, China, China
  • Mr Chongming Chen, State Key Laboratory of Multiphase Flow in Power Engineering, Xi'an Jiaotong University, China, China
  • Mr Yanmeng Gong, State Key Laboratory of Multiphase Flow in Power Engineering, Xi'an Jiaotong University, China, China
  • Mr Donghai Xu, State Key Laboratory of Multiphase Flow in Power Engineering, Xi'an Jiaotong University, China, China

    • Gasification of high moisture content biomass in supercritical water has been identified as a promising alternative system for producing renewable hydrogen. Many researchers have studied the reaction kinetics of supercritical water gasification of model compound. But real biomass which is of importance for large production of hydrogen has hardly been examined. In this paper, high moisture municipal sludge was studied by use of a batch reactor. The reaction was studied in the ranges of 673-723 K, 24-28 MPa and various reaction time 10-45 min. In order to predict the effect and show the principle of gasification of real organic wastes in supercritical water, artificial neural network is used to simulate the process. The reaction conditions and the elemental composition are selected as the input of the simulation. The simulation results of gas production rates fit well with the experimental ones. The artificial neural network can show the relation between output rate of gas composition and reaction conditions. The maximal deviation of the model is 34.33%. Artificial neural network provides a reliable method to predict the gas production composition of real organic wastes and thus the optimization of reaction conditions.