378960 Characterization and Modeling of Continuous Processing Unit Operations with Experimentally Measured Residence Time Distributions (RTD)

Wednesday, November 19, 2014
Galleria Exhibit Hall (Hilton Atlanta)
William E. Engisch Jr., Chemical & Biochemical Engineering, Rutgers University, Piscataway, NJ and Fernando J. Muzzio, Department of Chemical & Biochemical Engineering, Rutgers University, Piscataway, NJ

Residence time distributions (RTDs) are a valuable predictive tool in continuous processing systems including the continuous powder systems that can be used in pharmaceutical secondary manufacturing.  RTDs are commonly investigated in detail for fluid systems, while continuous powder unit operations are relatively unexplored.  The RTD of a system mathematically describes how a material travels inside the unit operations of a continuous processing system.  This has valuable in raw material traceability, sensing, and control.  Utilizing RTD it is possible to trace any upset or process change through the entire system, which can be used to improve quality.  Since the RTD is the pulse response of the system, it can also be used to define the frequency of sensing required to detect any instantaneous pulse perturbation.

This work focuses on the experimental trials used for measurement and determination of residence time distributions of all the processing equipment used in a direct compaction (DC) pharmaceutical line.  The DC process line consists of:  several feeders, a mill, a continuous blender, and a tablet press.  Each unit operations is investigated individually as well as in the overall line.  The RTDs are determined through a series of experiments where a tracer is added to the system or the individual unit operations in either a pulse or step-change and a NIR system is used to measure the resultant outlet stream.

The RTDs of each unit operation within the DC line are then programmed into a flowsheet modeling software (gPROMS), where disturbances to the system are simulated.  A statistical approach is taken to determine adequate sampling strategies for detecting various disturbances within the system.  Operational characteristic curves are generated for the various sampling strategies, which are then compared to one another as well as industry standards, such as USP <905>.

Extended Abstract: File Not Uploaded
See more of this Session: Poster Session: Pharmaceutical
See more of this Group/Topical: Food, Pharmaceutical & Bioengineering Division