Neural Network Based Modeling and Operational Optimization of Biomass Gasification Processes

Carbon neutral methanol from pulp mills towards full energy decarbonization: an inside perspective and critical review

“…The obtained PG can be used in various applications including thermal power generation or biofuel and biochemical syntheses. 131–134 However, due to some drawbacks, such as high tar (mixture of highly aromatic organic compounds 135 ) content 136 and low H 2 /CO ratio, 137,138 the raw PG is only suitable for thermal applications ( e.g. , lime kilns and boilers) and cannot be used directly for methanol synthesis.…”

“…The obtained PG can be used in various applications including thermal power generation or biofuel and biochemical syntheses. [131][132][133][134] However, due to some drawbacks, such as high tar (mixture of highly aromatic organic compounds 135 ) content 136 and low H 2 /CO ratio, 137,138 the raw PG is only suitable for thermal applications (e.g., lime kilns and boilers) and cannot be used directly for methanol synthesis. In fact, in a Biomass-to-Liquid (BtL) plants (Table 4), methanol production is performed by direct hydrogenation of syngas (R3, section 2.1), which requires specific stoichiometric ratios (R SR , eqn (1)) of around 2, 7,139 with higher values being beneficial for the kinetics.…”

Modelling and optimization of a pilot-scale entrained-flow gasifier using artificial neural networks

“…55 Souza et al modelled a biomass gasification process in a fluidized bed gasifier based on the concepts of ANN to correlate between the composition of the produced gas and the characteristics of different biomasses for several operating conditions. 56 Pandey et al developed an ANN-based modelling approach to estimate the low heating value of the gasification products. 57 In a study performed by Kallol Roy et.…”

Environmental Friendly Fluidized Bed Combustion of Solid Fuels: A Review About Local Scale Modeling of Char Heterogeneous Combustion

“…As it was discussed above, the formulation of a mechanistic model demands high efforts from the computational point of view, as well as the estimation of the properties of different solid fuels and the evaluation of the model parameters. This situation frequently results in a very simplified model with restricted applicability, as quoted by Bezerra de Souza et al [59]. Following the description made in [59], the Artificial Neural Networks (ANNs) are universal approximators that can be applied to various physical systems.…”

“…This situation frequently results in a very simplified model with restricted applicability, as quoted by Bezerra de Souza et al [59]. Following the description made in [59], the Artificial Neural Networks (ANNs) are universal approximators that can be applied to various physical systems. This technique allows to recognize strongly nonlinear relationships and to organize the information in a nonlinear mode in the context of empirical or hybrid modeling [60].The use of ANNs for modeling the solid fuel thermal treatments is currently a standard tool.…”