Statistical data treatment for residence time distribution studies in pharmaceutical manufacturing

Area of Science:

• Pharmaceutical Manufacturing
• Chemical Engineering
• Process Analytical Technology (PAT)

Background:

• Residence Time Distribution (RTD) is crucial for understanding powder dynamics in pharmaceutical manufacturing.
• Accurate RTD profiles are essential for predictive modeling and quality assurance.
• Significant noise in RTD measurements can hinder accurate analysis and modeling.

Purpose of the Study:

• To investigate the impact of noise on RTD measurements and applications.
• To evaluate different denoising methods for RTD profiles.
• To assess the effect of denoising on RTD-based applications like OOS analysis and modeling.

Main Methods:

• Simulated varying noise levels in RTD profiles using different tracers.
• Quantified the impact of noise using time and frequency averaging denoising methods.
• Evaluated Savitsky Golay filtering for its effectiveness in noise reduction.

Main Results:

• Savitsky Golay filtering proved effective for denoising RTD profiles across various noise levels.
• Key features of the RTD profile were preserved after denoising.
• Noise levels considered did not significantly impact RTD-based applications such as OOS analysis and RTD modeling.

Conclusions:

• Savitsky Golay filtering is a suitable method for handling noise in RTD measurements.
• Effective noise handling ensures the reliability of RTD-based applications in pharmaceutical manufacturing.
• The study provides practical insights into data treatment strategies for RTD studies.