Numerical Simulation as a Tool to Improve Subsurface Flow
89
Autotrophic bacteria are responsible for nitrification, which is modelled as a
two-step process. Nitrosomonas consume ammonia and produce nitrite,
Nitrobacter transform nitrite to nitrate. Both steps of the nitrification can only
occur when oxygen is available.
All microorganisms are assumed to be immobile and therefore they occur only
in the solid phase. For all microorganisms lysis represents the sum of all decay
and loss processes.
The total number of parameters in CW2D is 46 (26 kinetic parameters, 6
stoichiometric parameters, 8 composition parameters describing the nitrogen and
phosphorus content of the COD fractions, 2 parameters describing the oxygen
transfer between gaseous and aqueous phase and 4 parameters describing the
temperature dependency of the kinetic parameters). Additionally the material
parameters, and diffusion and adsorption parameters as used in HYDRUS-2D
have to be determined for each component. A detailed description of the processes
and parameters of CW2D is available (Langergraber 2000).
A sensitivity analysis shows the most influential parameters of CW2D
(Langergraber 2001). In general, one can summarise that the simulation results
show a great sensitivity to the van Genuchten model parameters describing the
unsaturated hydraulic properties of the main layer, the oxygen re-aeration rate and
the yield coefficients and lysis rates of the bacteria. The great influence of the
parameters of the unsaturated hydraulic property models on the reactive transport
simulation results gives the need for a good calibration of the flow model before
running reactive transport simulations with CW2D.
3 Results and Discussion
Simulation results are shown for a vertical flow pilot-scale constructed wetland
(PSCW) for wastewater treatment. Figure 2 shows the construction of the PSCW
that was based on the Austrian standards for vertical-flow constructed wetlands
for wastewater treatment (ONORM B 2505, 1997).
The PSCW had a surface area of 1 m 2 • The height of the main layer of the filter
bed was 60 cm. The main layer consisted of sand (gravel size 0/4 mm). An
intermediate layer of 10 cm with a gravel size of 4/8 mm prevented fine particles
from being washed out into the drainage layer. The effluent was collected in the
drainage layer (15 cm; gravel 16/32 cm) by means of drainage pipes.
The PSCW was loaded intermittently with wastewater. The daily loading rate
related to the standards was 40 I and the feeding interval was 6 h. Therefore a
single loading was 10 I. The pressure potential and the water content were
measured on-line at different depths (each lO cm in the main layer). Additionally,
the effluent flow rate was measured.
89
Autotrophic bacteria are responsible for nitrification, which is modelled as a
two-step process. Nitrosomonas consume ammonia and produce nitrite,
Nitrobacter transform nitrite to nitrate. Both steps of the nitrification can only
occur when oxygen is available.
All microorganisms are assumed to be immobile and therefore they occur only
in the solid phase. For all microorganisms lysis represents the sum of all decay
and loss processes.
The total number of parameters in CW2D is 46 (26 kinetic parameters, 6
stoichiometric parameters, 8 composition parameters describing the nitrogen and
phosphorus content of the COD fractions, 2 parameters describing the oxygen
transfer between gaseous and aqueous phase and 4 parameters describing the
temperature dependency of the kinetic parameters). Additionally the material
parameters, and diffusion and adsorption parameters as used in HYDRUS-2D
have to be determined for each component. A detailed description of the processes
and parameters of CW2D is available (Langergraber 2000).
A sensitivity analysis shows the most influential parameters of CW2D
(Langergraber 2001). In general, one can summarise that the simulation results
show a great sensitivity to the van Genuchten model parameters describing the
unsaturated hydraulic properties of the main layer, the oxygen re-aeration rate and
the yield coefficients and lysis rates of the bacteria. The great influence of the
parameters of the unsaturated hydraulic property models on the reactive transport
simulation results gives the need for a good calibration of the flow model before
running reactive transport simulations with CW2D.
3 Results and Discussion
Simulation results are shown for a vertical flow pilot-scale constructed wetland
(PSCW) for wastewater treatment. Figure 2 shows the construction of the PSCW
that was based on the Austrian standards for vertical-flow constructed wetlands
for wastewater treatment (ONORM B 2505, 1997).
The PSCW had a surface area of 1 m 2 • The height of the main layer of the filter
bed was 60 cm. The main layer consisted of sand (gravel size 0/4 mm). An
intermediate layer of 10 cm with a gravel size of 4/8 mm prevented fine particles
from being washed out into the drainage layer. The effluent was collected in the
drainage layer (15 cm; gravel 16/32 cm) by means of drainage pipes.
The PSCW was loaded intermittently with wastewater. The daily loading rate
related to the standards was 40 I and the feeding interval was 6 h. Therefore a
single loading was 10 I. The pressure potential and the water content were
measured on-line at different depths (each lO cm in the main layer). Additionally,
the effluent flow rate was measured.