Phenomenological Over-Parameterization of the Triple-Fitting-Parameter Diffusion Models in Evaluation of Gas Diffusion in Coal

• Main Authors: Jie Zang, Kai Wang, Ang Liu
• Format: Article
• Language: English
• Published: MDPI AG 2019-04-01
• Series: Processes
• Subjects: phenomenological over-parameterization; triple-fitting-parameters; coal diffusivity; gas diffusion; model
• Online Access: https://www.mdpi.com/2227-9717/7/4/241

Gas diffusivities of coal are not measured directly but are normally regressed by fitting mathematical diffusion models to fractional sorption data measured in sorption experiments. This paper firstly measured three fractional adsorption curves at various equilibrium pressures with the manometric method. The measured fractional adsorption curves were then modeled with one single-fitting-parameter (SFP) model and three triple-fitting-parameter (TFP) models. The modeling results showed that the TFP models were phenomenologically over-parameterized due to the usage of three fitting parameters, which may be excessive for curve fitting. The phenomenological over-parameterization resulted in multiple pressure-dependences of gas diffusivity for the TFP models. The TFP models should thus be carefully used. On the other hand, the dual-fitting-parameter (DFP) models also have excellent performance in curve fitting and can produce interpretable modeling results. The DFP models can be used as an alternative to the TFP model in the future.