Modeling and simulation of continuous powder blending applied to a continuous direct compression process.

Area of Science:

• Pharmaceutical Engineering
• Chemical Process Modeling
• Continuous Manufacturing

Background:

• Continuous manufacturing is gaining traction in the pharmaceutical sector.
• Powder blending is a critical unit operation for producing solid dosage forms like tablets.
• Optimizing powder blending is essential for ensuring product quality and process efficiency.

Purpose of the Study:

• To develop and validate a modeling methodology for describing residence time distribution (RTD) and blend uniformity in continuous powder blending systems.
• To assess the applicability of axial and radial tanks-in-series models for simulating a commercial powder blending process.
• To provide a predictive tool for optimizing continuous powder blending operations.

Main Methods:

• A tanks-in-series flowsheet modeling methodology was developed, incorporating both axial and radial tanks.
• Experimental data from impulse tests on a commercial powder blending system (GEA Pharma Systems) were used to generate residence time distributions (RTDs).
• Parameter estimation techniques were employed to fit the model to experimental RTD data, determining the number of axial tanks and validating model accuracy.

Main Results:

• The developed tanks-in-series modeling methodology demonstrated a good fit with experimental data, with weighted residuals below the chi-squared value at 95% confidence.
• In-silico impulse tests confirmed the model's ability to accurately describe the RTD behavior of the powder blenders.
• The simulations successfully predicted blend uniformity, with outputs falling within the experimentally observed variance.

Conclusions:

• The tanks-in-series modeling methodology, utilizing both axial and radial tanks, is a robust approach for characterizing continuous powder blending systems.
• This modeling approach provides valuable insights into RTD and blend uniformity, crucial for process development and control in pharmaceutical manufacturing.
• The validated model can be used for in-silico predictions, aiding in the optimization and scale-up of continuous powder blending operations.