Two unstructured segregated kinetic models to describe the anaerobic digest
ion of livestock manure are developed and experimental batch data obtained
from beef cattle in a 2.0-1 work volume digestor fitted to both proposed ki
netic models to obtain the values of the parameters. The results obtained b
y fitting show that the second model proposed has both statistical and phys
ical meaning in the parameter values obtained. The model takes into account
a simplified reaction scheme formed by six reactions. Several simplificati
ons have been made (lumping, pseudo-steady state for one compound, first or
der kinetics, etc.) yielding four key compounds to be analysed and fitted t
o the model as production-rates expressions (total biomass, chemical oxygen
demand (COD), volatile acids, and methane). The model considers three main
stages in the process: enzymatic hydrolysis of the waste, growth of 'aceto
genic' microorganisms (production of acids nongrowth associated), and growt
h of 'methanogenic' microorganisms associated with methane production; the
two last processes are accompanied by substrate consumption for maintenance
. A non-linear multiple-response regression technique coupled to a fourth-o
rder Runge-Kutta algorithm has been employed to obtain the values of the te
n parameters. The model is able to reproduce the experimental data obtained
for beef manure anaerobic digestion with more accuracy than experimental e
rror. (C) 1999 Elsevier Science Inc. All rights reserved.