ISSN Online: 2377-424X
ISBN CD: 1-56700-226-9
ISBN Online: 1-56700-225-0
International Heat Transfer Conference 13
INVERSE DETERMINATION OF BIOCHEMICAL MODEL PARAMETERS IN POROUS MEDIUM CONVECTION WITH BIOCHEMICAL HEATING
Sinopsis
The processes involved in biological reactors, such as in composting reactors for instance, are essentially non-isothermal, controlled by internal heat generation from microbial oxidation. This activity can be described adequately by a biochemical model of the Monod type. The inverse problem approach with adjoint equations can be adapted to this problem to achieve a solution for the determination of the maximum specific growth rate constant from internal temperature measurements.
The volumetric source is assumed proportional to the rate of consumption of a substrate by a biomass. A formulation based on the conjugate gradient method with adjoint equations is given for an arbitrary domain in two dimensions, with a Monod model to represent the biochemical activity. The direct, sensitivity and adjoint sets of equations are derived for a Boussinesq fluid, with convection-diffusion of the substrate within the porous matrix. Numerical calculations by control volumes are carried out for Rayleigh and Lewis numbers involved in a typical problem under realistic temperature and concentration boundary conditions. The influence of noisy input data and the issue of regularization are considered also.
The volumetric source is assumed proportional to the rate of consumption of a substrate by a biomass. A formulation based on the conjugate gradient method with adjoint equations is given for an arbitrary domain in two dimensions, with a Monod model to represent the biochemical activity. The direct, sensitivity and adjoint sets of equations are derived for a Boussinesq fluid, with convection-diffusion of the substrate within the porous matrix. Numerical calculations by control volumes are carried out for Rayleigh and Lewis numbers involved in a typical problem under realistic temperature and concentration boundary conditions. The influence of noisy input data and the issue of regularization are considered also.