Model Predictive Feeding Rate Control in Conventional and Single-use Lab-scale Bioreactors: A Study on Practical Application

Abstract

A developed solution for fed-batch process modeling and model predictive control (MPC), facilitating good manufacturing practice (GMP) based on process elaboration, control, and validation, is presented in the paper. The step-by-step evolution of the socalled "golden batch" optimal biomass growth profile and its control during the process is demonstrated. The case study of an advanced fed-batch control was performed on the recombinant E. coli BL21 lab-scale (5.4 L) biomass production process using the conventional stirred tank glass reactor. Additionally, a test experiment for control reproducibility and applicability assessment of the proposed approach was carried out in a single-use stirred tank reactor (5.7 L). Four sequentially performed experiments are demonstrated as an example for desirable feeding profile evolution for E. coli BL21 biomass production in a glucose-limited fed-batch process. Under different initial biomass and glucose conditions, as well as for different reference feeding profiles selected in the explorative experiments, good tracking quality of preset reference trajectories by the MPC system has been demonstrated. Estimated and experimentally measured biomass mean deviations from the preset reference value at the end of the processes were 4.6 and 3.8 %, respectively. Biomass concentration of 93.6 g L -1 (at 24 h) was reached in the most productive run. Better process controllability and safer process run, in terms of avoiding culture overfeeding but still maintaining a sufficiently high growth rate, was suggested for the process with biomass yield of 79.8 g L -1 (at 24 h). Practical recommendations on the approach application and adaptation for fed-batch cultures of interest are provided.

Application of In-Situ and Soft-Sensors for Estimation of Recombinant P. pastoris GS115 Biomass Concentration: A Case Analysis of HBcAg (Mut+) and HBsAg (MutS) Production Processes under Varying Conditions

“…The data exchange of the process and control variables between the control algorithms (Matlab) and SCADA (programmed in the software platform of PcVue Solutions, Ltd., Sèvres, France) was implemented every 1 s through an OPC server. Detailed information on the reactor vessel and control system configuration can be found in the research published earlier [ 52 ].…”

HBsAg Production in Methanol Controlled <i>P. pastoris</i> GS115 Mut^S Bioreactor Process

“…Recombinant Pichia pastoris GS115 strain (Mut S phenotype) carrying a gene for HBsAg synthesis was cultivated in a series of experiments in the 5 L fully automated benchtop fermenter bioreactor system (Biotehniskais Centrs, EDF-5.4/BIO-4). Detailed information on the reactor vessel and control system configuration can be found in earlier research [21]. The experimental plan included cultivation under two different cultivation protocols, one proposed by Invitrogen corporation [8] and second by Gurramkonda et al [3].…”

The Application of Adaptive Model Predictive Control for Fed-Batch Escherichia coli BL21 (DE3) Cultivation and Biosynthesis of Recombinant Proteins

“…To implement run-to-run optimization, a series of five cultivation experiments were carried out. In the first experiment, to model the target trajectories for the E. coli cultivation process (without biosynthesis of recombinant protein), the model coefficient values given in the article by O. Grigs et al [32] for the E. coli BL21 strain were used. Since a different correlation equation for biomass concentration versus OD was applied in the present work, the model parameters, e.g., Y xs , σ max , K Xmax , and X crit were adjusted.…”