Design of Experiments 2019

Design of Experiments 2019 is a course dedicated to the Design of Robust Chemical Processes.

Topics

• Statistical Design of Experiments (DoE)
  • Responses
  • Requirements for a successful DoE
  • Meaningful use of categorical factors
  • Scale independent and scale dependent factors
  • Ranges, number of levels
  • Factor ranking; factors to be fixed, factors to be varied
  • Randomization, replication, centerpoints, blocking
• Introduction
  • Statistics for chemical process R&D: friend not foe
  • Why DoE?
  • Univariate experimentation limitations
  • Synergy between statistical, kinetic and engineering models
  • Design of experiments (DoE), a multivariate method to investigate multivariate processes
• Analyzing DoE screening investigations
  • Key statistical concepts
  • Objectives
  • Practical vs. Statistical significance
  • The value of redundancy, the power of visualization tools
  • Case studies: chemical reactions, API crystallizations
• Screening the Experimental Space
  • Full factorial designs
  • DoE commercial software platforms
  • Plackett-Burman designs
  • Fractional factorial designs: advantages and challenges
  • Chemical reaction, and workup: together or separate?
  • Design efficiency, design resolution, statistical power of a design
  • Using DoE to set raw material specifications
  • Design augmentation; practical strategies for cost-effective screening
• DoE and Quality by Design (QbD)
  • Critical Quality Attributes and Critical Process Parameters
  • The concept of design space; risk calcualtions using the DoE model
  • DoE for process robustness assessment, and for process validation; process capability indices
• Response Surface Methodology (RSM)
  • RSM design options
  • Types of "optimization"
  • Model verification experiments
  • Analysis of RSM designs, model manipulation
  • Case studies: chemical reactions, API crystallizations
• Final Review
  • Round table discussions; practical tips, references, software demonstration
• "Advanced Topics" (time permitting)
  • DoE for process troubleshooting
  • DoE and Principal Component Analyisis (PCA)
  • Mixture designs
  • DoE investigating scale-dependent and scale-independent factors