Control of a Thiobacillus denitrificans bioreactor using machine vision

Area of Science:

• Biotechnology and Bioprocess Engineering
• Environmental Microbiology
• Industrial Microbiology

Background:

• Bioreactors are crucial for microbial processes, but challenges like biomass monitoring and byproduct formation (elemental sulfur) can lead to reactor upsets.
• Thiobacillus denitrificans bioreactors are susceptible to sulfide accumulation, causing colloidal elemental sulfur precipitation and process inefficiency.
• Effective real-time monitoring and control are essential for optimizing bioreactor performance and preventing operational issues.

Purpose of the Study:

• To develop and implement a PC-based machine vision system for continuous monitoring of biomass concentration in a Thiobacillus denitrificans bioreactor.
• To utilize the vision system for real-time detection and control of colloidal elemental sulfur production, a key indicator of reactor upset.
• To integrate intelligent process control by adjusting hydrogen sulfide feed rates based on vision system feedback.

Main Methods:

• A machine vision system employing a video camera with varied background lighting was established to capture digital images of the bioreactor.
• Analysis of red, green, and blue (RGB) intensity components within specific image regions was performed to track biomass concentration and detect elemental sulfur formation.
• A stepper motor-driven pressure regulator was utilized for automated adjustment of the hydrogen sulfide feed flow rate, controlled by the vision system's measurements.

Main Results:

• The machine vision system successfully monitored biomass concentration changes through color variations in the digital images.
• The ratio of red to blue intensity components proved effective in identifying the formation of elemental sulfur precipitates, indicating a reactor upset condition.
• Intelligent process control was achieved by dynamically altering the hydrogen sulfide feed flow rate based on real-time vision system data, mitigating upset conditions.

Conclusions:

• A PC-based machine vision system offers a viable and effective method for real-time monitoring of biomass and control of elemental sulfur precipitation in Thiobacillus denitrificans bioreactors.
• Color analysis of digital images, specifically the RGB intensity components and their ratios, provides a reliable indicator for biomass concentration and reactor upset detection.
• The integration of machine vision with automated process control enables intelligent adjustments to bioreactor operation, enhancing stability and efficiency.