Development of robust modelling methods for a quick evaluation of biomass pyrolysis performance and its transient behaviour at a particle level

Student thesis: Doctoral Thesis

Abstract

Biomass is organic matter that can be used as an energy resource by means of conversion technologies, such as pyrolysis. To assess pyrolysis performance of various biomass materials under different operating conditions and improve the process efficiency of biomass pyrolysis, the development of accurate, fast and robust modelling methods is desirable.The main objective of this thesis is to develop and assess modelling methods to obtain a deeper insight into the physico-chemical processes affecting a biomass subjected to pyrolysis, more specifically its (1) kinetics, and (2) mass and (3) heat transfer at a particle level.The kinetics of biomass pyrolysis have been studied for biomass in the thermally thin regime, and a novel method to obtain the kinetics parameters has been developed. This method can successfully provide a rapid and accurate estimation of the relative contributions of cellulose, hemicellulose, and lignin to the volatile yield, as well as their kinetic parameters.The method offers a simple way to obtain the kinetics parameters directly from thermogravimetric data and saves computing time by providing sensible initial values and bounds to the parameters. Wheat straw pellets have been used to study biomass pyrolysis in the thermally thick regime. The heat and mass transfer mechanisms that take place inside the pellet during pyrolysis have been analized using a single particle model.As a result, the changes inside a biomass pellet can be predicted at any given inner position and time during pyrolysis, and the inner gradients can be observed. It is concluded that the inner temperature distribution of the pellet depends mainly on the properties of the solid phase, but the final product distribution is also linked to the flux of the vapour phase, due to potential additional pyrolysis reactions that could take place if the volatiles generated are not quickly removed from the particle. All the models implemented are validated against experimental data.
Date of Award13 Mar 2020
Original languageEnglish
Awarding Institution
  • University Of Strathclyde
SponsorsUniversity of Strathclyde
SupervisorJun Li (Supervisor) & Leo Lue (Supervisor)

Cite this

'