The dimensional design of a laboratory-scale fluidized bed gasifier using machine learning based on a kinetic method

Kartal F., Özveren U.

Energy Conversion and Management, cilt.269, 2022 (SCI-Expanded) identifier identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 269
  • Basım Tarihi: 2022
  • Doi Numarası: 10.1016/j.enconman.2022.116183
  • Dergi Adı: Energy Conversion and Management
  • Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Academic Search Premier, PASCAL, Aerospace Database, Applied Science & Technology Source, CAB Abstracts, Communication Abstracts, Computer & Applied Sciences, Environment Index, INSPEC, Pollution Abstracts, Veterinary Science Database, Civil Engineering Abstracts
  • Anahtar Kelimeler: Kinetic modeling, Gasification, Circulating fluidized bed gasifier, Aspen Plus, ANN, ARTIFICIAL NEURAL-NETWORK, BIOMASS GASIFICATION, COAL-GASIFICATION, DYNAMICS SIMULATION, MODEL, CFD, CONVERSION, PARTICLE, REACTOR, FLOW
  • Marmara Üniversitesi Adresli: Evet

Özet

Gasification provides various environmental and technological advantages, and the efficiency of the gasification system is affected by several factors, including the kind of fuel and gasification agent used, the gasifier's length and diameter, the operating pressure and temperature, etc. Experimental optimization approaches are more realistic, but they are time demanding; also, a reactor operating at high temperatures and pressures could be dangerous and expensive. Thus, researchers use a variety of modeling techniques, including the process simulators. Additionally, artificial neural network (ANN) as a machine learning approach, which is one of the process modeling methods, is a remarkable approach, and several papers have been published in which it has been utilized in combination with other modeling techniques. On the other hand, a combined process simulator/ANN model that considers gasifier design/operational parameters for the kinetic modeling of gasification process has not been reported. In this study, after kinetic modeling and validation of seven different circulating fluidized gasifiers using Aspen Plus, parametric studies were performed. Parametric analysis was used to examine the impacts of gasifier diameter, length, gasifier temperature, air/fuel ratio, and fuel type, and a dataset was created for ANN training. The syngas composition and thermal value were predicted using the ANN model. Therefore, a model was developed that takes into consideration both design and operating variables. The investigations revealed that heterogeneous reactions were the most critical factor in defining syngas characteristics. Although design factors have a considerable impact on syngas characteristics, the gasifier temperature is a key factor in the whole process. Furthermore, the ANN model estimates syngas specifications with great accuracy (R2 > 0.99 and MAPE < 3%) based on fuel attributes and gasifier design/operating parameters. Hence, ANN models can be used to analyze the effectiveness of systems including a complex combination of reactions and thermochemical processes.