MATLAB-empowered brightness defect prediction system in pulp processing bleaching stage: An empirical modelling approach

Pulp quality significantly affects paper products, necessitating a balance between brightness, strength, and environmental sustainability. The bleaching process, which includes delignification and brightening stages, is crucial for achieving high pulp brightness. Current pulp bleaching research emphasizes optimizing processes and developing predictive models for better quality control, yet real-time pulp brightness monitoring remains a challenge. This research developed a MATLAB program to predict pulp brightness and consistency in real-time during bleaching, conducted entirely without incurring any financial costs. Empirical models for predicting pulp consistency at the extraction-oxidative-peroxide (EOP) and D1 stages were created using second-order polynomial equations, incorporating production rate and inlet pressure as variables. Brightness increment correlations were formulated based on temperature, chemical flow rate, residence time, pH, and inlet pressure, with specific models for the preceding chlorine dioxide (DA), EOP, and second chlorine dioxide (D1) stages. Data normalization ensured efficient processing by standardizing parameter scales. Results showed relationships between brightness increment and parameters for each sub-stage such as DA is linear for pH, quadratic for chlorine dioxide (ClO2) flow rate, cubic for temperature and residence time; EOP is linear for hydrogen peroxide (H2O2) flow rate, quadratic for temperature, cubic for inlet pressure, residence time, and pH; D1 is linear for pH, quadratic for ClO2 flow rate and residence time, cubic for inlet pressure and temperature. The coefficient of determination (R2) for the DA, EOP, and D1 sub-stages are 0.85313, 0.86526, and 0.86322, respectively. Parameters with the highest contributions in each stage were identified, such as inlet pressure in the D1 substage yielding the highest brightness gain. This system offers an alternative approach for analyzing pulp quality issues and is adaptable to future mill operational needs.