Instrumentação e controle de biorreatores utilizando impressão 3D: modelagem e avaliação de processos com aeração e formação de espuma

Abstract

The study developed and evaluated solutions aimed at improving the operational efficiency of bioreactors, with emphasis on mass transfer and foam formation. Additively manufactured devices produced in PLA, PETG, and photopolymer resins were designed in SOLIDWORKS and integrated with ESP32 and Arduino UNO microcontrollers for automation and process control. This integration enabled continuous data acquisition and real-time control of sensors and actuators for differential pressure, liquid and foam level, gas flow rate, biomass, temperature, pH, and dissolved oxygen. The determination of kLa was refined through a sigmoidal model with automated plateau detection and identification of dynamic parameters. The model was subsequently integrated into a temporal convolutional network, which provided real-time segmentation and computationally efficient estimates suitable for data processing and online applications. The investigation of nonconventional gas dispersion methods identified the gas induction impeller and the Venturi hydroejector as efficient alternatives. The hydroejector (3 vvm) reached a maximum kLa of 51.2 ± 0.8 h⁻¹, exceeding the stirred tank bioreactor operated (400 rpm and 3 vvm) by 26 %, with equivalent energy efficiency. In ethanol stripping during Saccharomyces cerevisiae cultures, the hydroejector produced an average removal coefficient (ke) of 0.0235 ± 0.0011 h⁻¹ at 34°C, accelerated substrate consumption, and partially mitigated fermentative inhibition. The gas recovery system, integrating condensation and adsorption in silica columns, retained more than 95 % of the entrained ethanol. These results were consistent with the kinetic modeling performed for conditions with and without stripping, which confirmed a higher specific consumption rate and reduced ethanol accumulation when the Venturi device was active. In aerobic cultivations with Bacillus subtilis, the evaluation of antifoam additives showed that appropriate supplementation can increase productivity. One formulation increased the final surfactin concentration from 1.12 ± 0.04 g/L to 2.04 ± 0.06 g/L without impairing cell growth. The cascade control strategy based on the interpolated kLa gradient enabled coordinated adjustment of aeration and agitation to maintain stable dissolved oxygen, advance surfactin production, and reduce energy consumption relative to fixed-operation strategies. The kinetic model for Bacillus subtilis, developed from two batch cultivations under controlled pH, provided parameters suitable for semi-continuous simulations. These parameters supported the formulation of an optimal fed-batch strategy in an internal-circulation airlift bioreactor, predicting 3.10 g/L surfactin with a volumetric productivity of 0.254 g/L/h over 61 hours. Therefore, the integration of additive manufacturing, electronic instrumentation, and analytical methods for mass transfer provides a structured basis for enhancing bioreactor performance, incorporating automatic foam and dissolved oxygen control, online kLa estimation, and gas-dispersion and stripping strategies applicable to process intensification in biochemical systems.