Model reliability and parameter estimability of gasification dynamic models

Gasification plays an important role for converting coal and biomass to syngas. These product gases can be applied to many processes, for example, power generation or methane production. There are a number of the mathematic models that can predict the composition of the product gas from the different type of gasification processes and feedstock, but these models are not compact and can be employed for just the specific type of gasification. In many cases, the collected experimental data are not sufficient to reliably identify the parameter estimates. In this case, the identification of all parameters without distinction reduces dramatically the prediction capabilities of the mathematical models and potentially to unreliable mathematical models. The estimablility approach was proven to be an effective way to improve the model prediction compatibility using the available experimental data. This approach was successfully implemented to gasification mathematical models and provides a methodology that helps exploit optimally the available experimental measurements. If the experimental data are not sufficient, additional experiments could be designed to improve the estimability potential of a specific parameter. As a demonstration of the methodology, the estimability framework was implemented and validated using the biomass gasification process. The results show that this method can provide the most influential parameter according to the different cut-off value, so the model parameters are able to be validated very effectively. This modified dynamic model has the potential to predict the gas composition produced from the biomass gasification process.