Bioprocess control utilising alternative specific rates

Abstract

To a considerable degree biotechnological upstream processes rely on offline sampling, feed forward as well as manual feed control, as no platform technologies for robust and transferable bioprocess monitoring and -control are available yet. Poor data quality and quantity impedes process automation as well as consistent product quality and safety. As a result controlling approaches rarely look beyond traditional biomass growth rate control, neglecting opportunities posed by alternative specific rates. This study employed an elemental balancing soft sensor approach extended by simple yet adaptive error estimation for bioprocess monitoring. Compared to an existing approach, online data obtained was greatly improved, as RMSE of formerly 22% was decreased to 11%. This enabled sensitive gross error detection and bioprocess control based on specific rates. With a solid monitoring basis various forms of specific rate control and its impact on product amounts yielded in an E. coli inclusion body process and a novel form of nitrogen control were investigated. In contrast to growth rate control, control of specific substrate rate provided steady feed rate set points during induction phase, as it was found to be uninfluenced by changes in cell metabolism. In an effort to account for changes in cell metabolism growth rate control was substituted with specific carbon emission rate control during induction phase, which improved specific titer (2.58 mg/g) compared to the reference process (2.23 mg/g), while still being inferior to specific substrate rate control (3.31 mg/g). The results arm outperformance of specific substrate rate control over growth rate control and suggest specific carbon emission rate control as means for metabolism directed feed termination.