Numerical model estimation of biomethane production using an anaerobic CSTR: model formulation, parameter estimation and uncertainty/sensitivity analysis

Hatem Yazidi*, Geraint Bevan, Joseph V. Thanikal, Ole Pahl, Colin Hunter

*Corresponding author for this work

Research output: Contribution to journalArticle

41 Downloads (Pure)

Abstract

In this paper, an innovative complex numerical model is proposed for simulating the anaerobic biogas production potential of organic waste materials both for full-scale anaerobic plant design and operation decisions and for laboratory and pilot scale co-digestion research. The model facilitates, in particular, the understanding of co-digestion and mixture effects by application of uncertainty and global sensitivity analysis. This allows multi-dimensional parameter analysis so that uncertainties and the main sensitivities can be identified among the model parameters, with a special focus on those leading to digester failure. The initial application of the complex model to ongoing lab-scale anaerobic co-digestion processes revealed that the hydrolysis and acidogenesis phases are the most affecting steps of the methane production. In particular, the following parameters have been found to contribute the most to the variance of the complex model’s estimate of methane production: polymer hydrolysis rate; specific acidogens maximum growth rate; saturation constant for acidogens, the specific acetoclastic methanogens maximum growth rate; saturation constant for acetoclastic methanogens; and the gas-liquid mass transfer coefficient for methane.
Original languageEnglish
Pages (from-to)58-67
Number of pages10
JournalARPN Journal of Agricultural and Biological Science
Volume13
Issue number6
Publication statusPublished - 30 Jun 2018

Keywords

  • anaerobic co-digestion
  • statistical sampling
  • uncertainty analysis
  • global sensitivity analysis

Fingerprint Dive into the research topics of 'Numerical model estimation of biomethane production using an anaerobic CSTR: model formulation, parameter estimation and uncertainty/sensitivity analysis'. Together they form a unique fingerprint.

  • Cite this