Determination of kinetic constants by using transient temperature data from continuous stirred tank reactors

Tapio Salmi*, Pasi Tolvanen, Kari Eränen, Johan Wärnå, Sébastien Leveneur, Heikki Haario

*Corresponding author for this work

Research output: Contribution to journalArticleScientificpeer-review

Abstract

Mathematical modelling of transient states of continuous stirred tank reactors was considered for the case of homogeneous liquid-phase reactions. A simple and very precise kinetic measurement strategy was proposed: registration of the reactor temperature during the start-up of the reaction process. The dynamic reactor model consisted of mass and energy balances written in dimensionless forms to enable general conclusions and easy computer implementation. Model simulations were carried out by using a numerical solver for stiff differential equations and a reaction invariant was revealed by theoretical analysis. A series of transient laboratory-scale experiments for a strongly exothermic reaction, a reaction between sodium thiosulphate and hydrogen peroxide was conducted and the kinetic parameters, the pre-exponential factors and activation energies were estimated. The data fitting was further improved by incorporating the decomposition of hydrogen peroxide as a side reaction. The estimated kinetic parameters were investigated further by sensitivity analysis and Markov-Chain-Monte-Carlo-Method (MCMC) to confirm their reliability. The proposed method is applicable for relatively simple systems with a measurable heat effect. For complex multireaction systems the method should be completed with chemical analysis.

Original languageEnglish
Article number117164
JournalChemical Engineering Science
Volume248
Early online date6 Oct 2021
DOIs
Publication statusE-pub ahead of print - 6 Oct 2021
MoE publication typeA1 Journal article-refereed

Keywords

  • Adiabatic CSTR
  • Homogeneous reactions
  • Rate constants
  • Thermal effect

Fingerprint

Dive into the research topics of 'Determination of kinetic constants by using transient temperature data from continuous stirred tank reactors'. Together they form a unique fingerprint.

Cite this