(460d) Design Space Identification Via Multi-Parametric Programming | AIChE

(460d) Design Space Identification Via Multi-Parametric Programming

Authors 

Sachio, S. - Presenter, Imperial College London
Papathanasiou, M., Imperial College London
Kontoravdi, C., Imperial College London
As demand for biopharmaceuticals rises and there is a drive towards sustainable processes, now biopharmaceutical manufacturing needs to meet multiple competing key performance indicators (KPIs), including sustainability, efficiency, and quality. Biopharmaceutical manufacturing is challenged by high manufacturing costs and large amount of waste (Nasr et al., 2017). The complex nature of bioprocesses and measurement unavailability are some of the key challenges of process control. This results in trial-and-error experiments in traditional practices which may be costly and inefficient (Narayanan et al., 2020). Quality by Design (QbD) promotes systematic decision making and deep understanding in pharmaceutical development and manufacturing (Rathore and Winkle, 2009). The key is to build product quality from the start. However, optimal design and online control in biopharmaceutical manufacturing still faces a number of unsolved challenges (Papathanasiou and Kontoravdi, 2020). In this work, we are proposing a new methodology for mapping the design space via multi-parametric programming to accelerate process design and control.

We focus on the capture step of monoclonal antibody (mAb) manufacturing using the twin column protein A chromatography model presented by Steinebach et al. (2016). The model comprises 10 equations. It uses a two-site isotherm to describe the adsorption mechanism, lumped driving force mass transfer kinetics to describe transport in pores, and partial differential equations to describe the column mass balance. The model presented by Steinebach et al. (2016) was simulated in gPROMS® ModelBuilder. Sensitivity analysis was performed using MATLAB and the SobolGSA toolbox for the identification of the most significant variables and parameters affecting product purity, yield, and productivity for the design space identification. The results showed that column switching time, column dimensions, resin properties, feed concentration and flowrate had significant impact.

We used multi-parametric programming to map the design space of the capture process for a variation of design variables. One of the main advantages of this approach was that the design space is described by a set of affine functions which map the variables and parameters to the KPIs. This can accelerate the online system evaluation, optimisation, and control (Diangelakis et al., 2017).

Figure 1 illustrates the framework as followed in this work. First, the high fidelity model was constructed and validated. Then, design space identification via GSA and optimisation were carried out. To solve the multi-parametric programming problem the high-fidelity model was adapted to reduce computational complexity (Pistikopoulos et al., 2015). Finally, the multi-parametric approach was carried out using the reduced model. The results were assessed and compared in terms of accuracy and reliability by validation with the high-fidelity model.

References

Diangelakis, N. A., Burnak, B., Katz, J. & Pistikopoulos, E. N. (2017). Process design and control optimization: A simultaneous approach by multi-parametric programming. AIChE Journal, 63, 4827-4846.

Narayanan, H., Luna, M. F., von Stosch, M., Cruz Bournazou, M. N., Polotti, G., Morbidelli, M., Butte, A. & Sokolov, M. (2020). Bioprocessing in the Digital Age: The Role of Process Models. Biotechnol J, 15, e1900172.

Papathanasiou, M. M. & Kontoravdi, C. (2020). Engineering challenges in therapeutic protein product and process design. Current Opinion in Chemical Engineering, 27, 81-88.

Pistikopoulos, E. N., Diangelakis, N. A., Oberdieck, R., Papathanasiou, M. M., Nascu, I. & Sun, M. (2015). PAROC—An integrated framework and software platform for the optimisation and advanced model-based control of process systems. Chemical Engineering Science, 136, 115-138.

Rathore, A. S. & Winkle, H. (2009). Quality by design for biopharmaceuticals. Nat Biotechnol, 27, 26-34.

Steinebach, F., Angarita, M., Karst, D. J., Muller-Spath, T. & Morbidelli, M. (2016). Model based adaptive control of a continuous capture process for monoclonal antibodies production. J Chromatogr A, 1444, 50-6.