SIDCOP Research Group, University of Antioquia, Medellín, Colombia.
Biotransformación Research Group, University of Antioquia, Medellín, Colombia.
Bioprocess Biosyst Eng. 2019 Jun;42(6):1023-1037. doi: 10.1007/s00449-019-02102-z. Epub 2019 Mar 14.
Fed-batch production of poly(3-hydroxybutyrate-co-3-hydroxyvalerate) copolymer using vinasses-molasses mixture is carried out in this work by implementing different process systems engineering tools. Two fed-batch strategies are tested experimentally at 5 L scale, considering only offline information: (1) offline optimizing control and (2) exponential feeding. Application of these strategies showed that different feeding profiles result in different dynamic behaviour, influencing both, yield and biopolymer properties. As offline-based feeding strategies do not consider information of the culture status, they cannot deal with uncertainties. Therefore, a closed loop control strategy was implemented, which uses biomass and substrate information predicted online by soft-sensors. Results demonstrated the technical feasibility to produce biopolymer using a 75/25%vol. vinasses-molasses mixture. Successful implementation of the soft-sensor-based control strategy was evidenced at pilot plant scale, where sugar concentration was kept almost constant for 14 h, while obtaining the desired copolymer. Thus, proposed control strategy could be of interest at industrial-scale.
本文采用不同的过程系统工程工具,利用糖蜜-酒糟混合物进行分批补料生产聚(3-羟基丁酸酯-3-羟基戊酸酯)共聚物。在 5 L 规模上,通过实验测试了两种分批补料策略,仅考虑离线信息:(1)离线优化控制和(2)指数进料。应用这些策略表明,不同的进料曲线导致不同的动态行为,影响产率和生物聚合物的性能。由于基于离线的进料策略不考虑培养状态的信息,因此它们无法处理不确定性。因此,实施了闭环控制策略,该策略使用软传感器在线预测的生物量和基质信息。结果表明,使用 75/25%vol 的酒糟-糖蜜混合物生产生物聚合物具有技术可行性。在中试规模上成功实施了基于软传感器的控制策略,其中糖浓度在 14 小时内几乎保持恒定,同时获得了所需的共聚物。因此,所提出的控制策略在工业规模上可能具有吸引力。