(344c) Modeling of Residence Time Distribution of a Continuous Dry Granulation Tableting Line | AIChE

(344c) Modeling of Residence Time Distribution of a Continuous Dry Granulation Tableting Line


Martinetz, M. C. - Presenter, Research Center Pharmaceutical Engineering GmbH
Karttunen, A. P., University of Eastern Finland
Korhonen, O., University of Eastern Finland
Sacher, S., RCPE
Khinast, J. G., Research Center Pharmaceutical Engineering
Wahl, P. R., RCPE GmbH
In recent time academia and industry have shown the economic and process related benefits of continuous pharmaceutical manufacturing. Regulatory authorities support strongly this approach of processing as well. This trend leads to a need for advanced control strategies, which enable automated operation of integrated production lines. With the first continuous plants being implemented for continuous production of solid oral dosage forms, out-of-spec material discharge strategies and process control with the aim to ensure robust product quality and to save resources have to be developed. Malfunctions of process equipment, unexpected process events and start-up or shut-down phases can lead to out-of-spec material. Due to material holdup and back mixing effects in specific unit operations, the out-of-spec material distribution can spread over process time. Therefore, knowledge of the residence time distribution (RTD) is essential not only for batch definition and material tracking aspects. For the development of cost-effective discharge strategies a fundamental tracking of the material through the process based on RTD is required.

In this work a dry granulation line, as one of the common ways of solid dosage form manufacturing, was investigated. The line consists of a feeding, blending, dry granulation, pneumatic transport and tableting step. For the determination of the residence time distribution in the considered unit operations as well as in the entire line a step change of API was performed at different total throughputs. As a system response the concentration of API was measured via in-line NIR spectroscopy at the outlet of blender, granulation and tablet press. Additionally, samples of powder, granules and tablets were drawn and analyzed off-line by means of UV/Vis as a reference.

Based on the experimental RTD data a transfer function was calculated and the transfer behavior of single unit operations modeled. The RTDs for single unit operations were interlinked and the entire process simulated with customized models in the process simulation software gSolids. The selected approach allows the prediction of out-of-spec material amount and the discharge time for different throughputs. This work shows promising results for both the inter- and extrapolation of RTDs for various throughputs and comparison with validation runs of the entire line.