Modeling and optimization of acetic acid fermentation: A polynomial-based approach

• Development of black box models with a high interest for the industrial process.
• First time that typical operational variables used in industry have been correlated.
• Easy to use for optimizing the process under different criteria.
• Model predictions have been validated by a different modeling approach.

Vinegar production is a typical bioprocess in the scope of the agrifood industry. Its optimization requires careful modeling which has so far been addressed by using mainly unstructured first principles models. Because of the difficulties in obtaining these models, black box models, such as those used here, are becoming more frequently used. The polynomial models developed in this work, accurately reflect the effect of the major and typical operational variables used in industry for this process. Also, response surfaces were used to identify the optimum operating conditions with a view to maximizing the mean fermentation rate and productivity. The followed strategy has a huge industrial interest since yields a tool that does not only allow finding the best operational conditions depending on different criteria but also is useful for process control. As far as we know this is the first time that these variables have been correlated in this way.