Batch Fermentation /w pH Inhibition.
Instructor: Nam Sun Wang
Model parameters
hr-1
g/L
g biomass/g substrate
Product formation
mole H/g biomass
mole H/(hr
This organism is actually quite pH tolerant.
Simulate batch fermentation.
g/L
g/L
mole/L
ODEs
... biomass
... substrate
... product (which is directly related to H)
I.C.
Length of time to complete fermentation
ODEs with pH controlled at pH=7
... biomass
... substrate
... product
I.C.
Length of time to complete fermentation
Note that biomass and product concentrations closely match, because product formation is dominated by the growth-related term α
Without pH control, fermentation lasted
hr
Product productivity
mole/(L
With pH control, fermentation lasted
hr
Product productivity
mole/(L
Product productivity with pH control is ~5X higher than that without pH control.
Estimate Model Parameter from Dynamic Data.
True:
True value:
Initial guess:
Estimated value:
... Good!
Estimate 4 model parameters simultaneously.
True:
True value:
Initial guess:
Estimated value:
Estimated values of x, s, pH are good.
Predicted:
Estimate 5 model parameters simultaneously.
True:
True value:
Initial guess:
Estimated value:
Estimated values of x, s, pH are good.
Predicted:
Estimate all 7 model parameters simultaneously.
True:
True value:
Initial guess:
Estimated value:
The estimated values of x, s, pH are ok.
Predicted:
In general, μm and Yx are estimated well. The estimated pK1 value is adjusted from the initial guess toward the true value, but the estimated pK2 value does not change much from the initial guess, because the pH dynamic data do not stay around the pK2 range.